1. Microbiology and Infectious Disease

Roles for mycobacterial DinB2 in frameshift and substitution mutagenesis

  1. Pierre Dupuy
  2. Shreya Ghosh
  3. Allison Fay
  4. Oyindamola Adefisayo
  5. Richa Gupta
  6. Stewart Shuman  Is a corresponding author
  7. Michael S Glickman  Is a corresponding author
  1. Immunology Program, Sloan Kettering Institute, United States
  2. Molecular Biology Program, Sloan Kettering Institute, United States
  3. Immunology and Microbial Pathogenesis Graduate Program, Weill Cornell Graduate School, United States
https://doi.org/10.7554/eLife.83094
Share this article
Cite this article
  1. Pierre Dupuy
  2. Shreya Ghosh
  3. Allison Fay
  4. Oyindamola Adefisayo
  5. Richa Gupta
  6. Stewart Shuman
  7. Michael S Glickman
(2023)
Roles for mycobacterial DinB2 in frameshift and substitution mutagenesis
eLife 12:e83094.
https://doi.org/10.7554/eLife.83094
  2. Download BibTeX
  3. Download .RIS

Abstract

Translesion synthesis by translesion polymerases is a conserved mechanism of DNA damage tolerance. In bacteria, DinB enzymes are the widely distributed promutagenic translesion polymerases. The role of DinBs in mycobacterial mutagenesis was unclear until recent studies revealed a role for mycobacterial DinB1 in substitution and frameshift mutagenesis, overlapping with that of translesion polymerase DnaE2. Mycobacterium smegmatis encodes two additional DinBs (DinB2 and DinB3) and Mycobacterium tuberculosis encodes DinB2, but the roles of these polymerases in mycobacterial damage tolerance and mutagenesis is unknown. The biochemical properties of DinB2, including facile utilization of ribonucleotides and 8-oxo-guanine, suggest that DinB2 could be a promutagenic polymerase. Here, we examine the effects of DinB2 and DinB3 overexpression in mycobacterial cells. We demonstrate that DinB2 can drive diverse substitution mutations conferring antibiotic resistance. DinB2 induces frameshift mutations in homopolymeric sequences, both in vitro and in vivo. DinB2 switches from less to more mutagenic in the presence of manganese in vitro. This study indicates that DinB2 may contribute to mycobacterial mutagenesis and antibiotic resistance acquisition in combination with DinB1 and DnaE2.

Editor's evaluation

This important study uses a combination of compelling biochemical and genetic approaches to identify a highly mutagenic mycobacterial DNA polymerase, which drives a wide spectrum of mutations when overexpressed. The findings advance the understanding of mutagenesis in mycobacteria. The work will be of interest to bacteriologists working on mutation mechanisms and the emergence of drug resistance.

https://doi.org/10.7554/eLife.83094.sa0

Introduction

One of the primary mediators of chromosomal mutagenesis is the error-prone DNA damage tolerance pathway termed translesion synthesis (TLS) (Vaisman and Woodgate, 2017). TLS polymerases transiently replace the replicative DNA polymerase to traverse lesions that block replication. Because of the flexibility of their active site and lack of proofreading activity, these polymerases facilitate survival during DNA damage but are mutagenic. TLS has been extensively studied in Escherichia coli, which encodes two DNA damage-inducible polymerases: DinB (Pol IV) and UmuDC (Pol V) (Fuchs and Fujii, 2013; Fujii and Fuchs, 2020). DinB and UmuDC confer tolerance to multiple forms of DNA damage and mediate mutagenesis by incorporating both substitution and frameshift mutations. Whereas DinBs are ubiquitous in bacteria, the distribution of UmuDC is more restricted. Many bacteria, including mycobacteria, encode a second copy of the replicative DNA polymerase (Pol III), called DnaE2 (Cole et al., 1998; Erill et al., 2006), as a translesion polymerase.

The Mycobacterium genus, which includes the causative agent of tuberculosis (TB) Mycobacterium tuberculosis (Mtb) and several others pathogens, encodes DnaE2 as well as several DinB paralogs (Cole et al., 1998; Timinskas and Venclovas, 2019). DnaE2 confers UV tolerance and antibiotic resistance in Mtb through its mutagenic activity and plays a role in pathogenicity (Boshoff et al., 2003). DnaE2 was initially thought to be the only active TLS polymerase in mycobacteria. The role of DinBs in DNA damage tolerance and mutagenesis was unclear because initial studies of dinB deletion strains failed to identify a function of Mtb DinBs (Kana et al., 2010). Mycobacterial DinBs exemplify three phylogenetic subfamilies found in many actinobacteria: DinB1/DinX, DinB2/DinP, and DinB3/msDinB3 (Cole et al., 1998; Timinskas and Venclovas, 2019). In a recent study, we revealed that Mycobacterium smegmatis and Mtb DinB1 contribute to alkylation damage tolerance, antibiotic resistance though a characteristic mutagenic spectrum, and chromosome diversification through frameshift mutations in homo-oligonucleotide runs (Dupuy et al., 2022). Some of these DinB1 activities, particularly homopolymeric run frameshift mutagenesis, are redundant with DnaE2 during exposure to DNA damage, suggesting that DinB1 and DnaE2, both of which interact with the β-clamp, exert their overlapping activities at the replication fork.

M. smegmatis encodes two additional DinBs, DinB2 and DinB3, whereas Mtb encodes DinB2 but lacks DinB3. Biochemical studies of M. smegmatis DinB2 and DinB3 showed that both have polymerase activity (Ordonez et al., 2014). DinB2 is notable for its capacity to utilize ribonucleotides during templated DNA synthesis, an activity that is attributable to the absence of a polar filter (Johnson et al., 2019) and a permissive steric gate (Ordonez et al., 2014). Leucine 14 of DinB2 replaces the canonical aromatic amino acid which clashes with the 2’-OH of rNTPs and thereby confers selectivity to DNA polymerases. DinB2 also displays a metal-dependent mutagenic switch in which manganese supports efficient 8-oxoguanine utilization and nucleotide addition opposite 8-oxoguanine in the template (Ordonez and Shuman, 2014). These biochemical activities suggest that DinB2 is equipped to mediate a diverse range of mutation types in vivo. However, our prior data (Dupuy et al., 2022) indicate that DinB2 is expressed at very low levels in basal conditions and not induced by DNA damage, in contrast to DnaE2, which is induced ~100-fold by UV. Indeed, Patra et al., 2021 showed that DinB2 expression is actively repressed in M. smegmatis by the action of the TetR family repressor protein MSMEG_2294 encoded in an operon with the dinB2 gene. This explains why we did not detect any effects of deleting the dinB2 ORF on mutagenesis in vivo (Dupuy et al., 2022), because DinB2 is effectively absent under the conditions surveyed. Thus, an alternative approach is needed.

In this study, we used inducible overexpression to gauge whether and how DinB2 (and DinB3) can affect genomic integrity. We show that DinB2 and DinB3 are both promutagenic in vivo, with distinct mutagenic signatures. In addition, we highlight the strong ability of DinB2, but not DinB3, to incorporate frameshift mutations in both short and long homo-oligonucleotide runs, an activity that is enhanced by manganese in vitro. Finally, we find that manganese enhances the growth inhibitory effects of DinB2 overexpression, suggesting that the metal switch operates in vivo.

Results

Overexpression of DinB2 causes cell death through its polymerase activity

To examine the function of DinB2 and DinB3 in M. smegmatis, we expressed plasmid-borne copies of the dinB2 or dinB3 genes, encoding untagged or streptavidin-tagged (ST) versions of M. smegmatis DinB2 and DinB3, under the control of an anhydrotetracycline (ATc)-inducible promoter (tet promoter). DinB2 and DinB3 were detected by immunoblotting with anti-ST antibodies after inducer addition. Levels of DinB2 and DinB3 were similar 4 and 24 hr after ATc treatment (Figure 1—figure supplement 1A). Compared to the untreated condition, the levels of DinB2 and DinB3 were increased by between 5- and 15-fold, depending on ATc concentration (Figure 1A). Overexpression of tagged and untagged versions of DinB2, but not DinB3, caused a growth defect (Figure 1B and Figure 1—figure supplement 1B) and 10-fold loss of viability (Figure 1C) which was proportional to the concentration of inducer (Figure 1—figure supplement 1B–C). DinB2 overexpression also triggered the DNA damage response, as measured by RecA induction (Figure 1D). Despite no effect on growth, overexpression of DinB3 also induced the DNA damage response (Figure 1D).

Figure 1 with 2 supplements see all
Download asset Open asset
Overexpression of DinB2 causes cell death through its polymerase activity.

(A) Anti-streptavidin/RpoB immunoblots from indicated strains with indicated concentrations of inducer treatment (16 hr of treatment). Average and SEM of RpoB normalized band intensities (n=3, arbitrary units) are given below the image of a representative blot. (B) Growth and (C) viability of indicated strains in presence of 50 nM anhydrotetracycline (ATc). Note that the OD600 values in (B) are calculated values based on continuous dilution growth experiments (see Methods). (D) Anti-RecA/RpoB immunoblots from indicated strains with indicated times of inducer treatment (50 nM). Average of normalized band intensities, expressed relative to the empty vector strain, is given below the image of a representative blot. (E) Anti-streptavidin/RpoB immunoblots from indicated strains after 16 hr of inducer treatment (50 nM ATc). Average and SEM of normalized band intensities (n=3) are given below the image of a representative blot. (F) Growth of indicated strains in presence of 50 nM ATc. (G) Anti-RecA/RpoB immunoblots from indicated strains after 24 hr of inducer treatment. Average of normalized band intensities, expressed relatively to the empty vector strain, is given below the image of a representative blot. Empty = empty vector, tet = ATc-inducible promoter, DinB2=M. smegmatis DinB2, DinB3=M. smegmatis DinB3, ST=streptavidin tag, D107A=catalytically inactive M. smegmatis DinB2. Results shown are means (± SEM) of biological triplicates. Stars under the means mark a statistical difference compared to the empty vector reference strain (**, p<0.01; ***, p<0.001).

Figure 1—source data 1

Uncropped immunoblots Figure 1A, D, E and G.

https://cdn.elifesciences.org/articles/83094/elife-83094-fig1-data1-v1.zip

Our prior work showed that overexpression of DinB1 in M. smegmatis also induces cell death, probably by interacting with the replicative machinery and competing with the replicative DNA polymerase (Dupuy et al., 2022), a phenotype that was exacerbated by a polymerase dead active site mutation in DinB1. Unlike DinB1, neither DinB2 nor DinB3 have a predicted β clamp binding motif (Kana et al., 2010). To investigate the cause of the cell death induced by DinB2 overexpression, we expressed DinB2-D107A, which lacks polymerase activity (Ordonez et al., 2014). Whereas DinB2 and DinB2D107A were expressed at similar levels (Figure 1E), DinB2D107A did not arrest bacterial growth (Figure 1F and Figure 1—figure supplement 1D) or induce the DNA damage response (Figure 1G). These results indicate that, unlike DinB1, the toxic effect of DinB2 is due to its polymerase activity.

A study conducted in E. coli revealed that dinB overexpression toxicity is due to genomic incorporation and excision of 8-oxoguanines, leading to the formation of DNA double-strand breaks (Foti et al., 2012). This is reminiscent of the ability of DinB2 to utilize 8-oxoguanine for DNA synthesis in vitro (Ordonez and Shuman, 2014). To test if DinB2 overexpression causes cell death in vivo because of genomic incorporation of 8-oxoguanines, we measured the impact of mutTs and mutY/mutMs deletion on DinB2-dependant lethality. The mutT and mutY/mutM enzyme systems are involved in degrading free 8-oxoguanine nucleotides and excision of genomic 8-oxoguanines, respectively (Dupuy et al., 2020). We did not observe a significant impact of the absence of these anti-8-oxoguanine systems on the dinB2 overexpression growth defect (Figure 1—figure supplement 2A and C) or loss of viability (Figure 1—figure supplement 2B and D), indicating that DinB2-dependent lethality is not mainly due to genomic 8-oxoguanine incorporation.

DinB2 and DinB3 confer antibiotic resistance through a distinct mutagenic profile

Previous studies showed that mycobacterial DnaE2 and DinB1 confer antibiotic resistance by stimulating chromosomal substitution mutations with distinct mutation signatures (Boshoff et al., 2003; Dupuy et al., 2022). To investigate if DinB2 and DinB3 are similarly mutagenic, we measured the rifampicin resistance (rifR) frequency in M. smegmatis strains overexpressing DinB2 or DinB3. In the absence of inducer, strains carrying the empty vector, tet-dinB2 or tet-dinB3 had similar rifR frequencies. Although 16 hr of inducer treatment had no effect on the rifR frequency in the control strain, overexpression of DinB2 or DinB3 increased the rifR frequency by sixfold, compared to –ATc condition (Figure 2A). A similar effect was observed with an ST version of DinB2 but not DinB2D107A-ST(Figure 2B), showing that the polymerase activity of DinB2 is necessary to confer antibiotic resistance. DinB2 and DinB3 overexpression still enhanced rifR frequency in the ΔrecA and ΔdnaE2 strains (Figure 2—figure supplement 1), indicating that the antibiotic resistance conferred by DinB2 is not an indirect consequence of the activation of the DNA damage response or the induction of the DNA damage-inducible error-prone DnaE2 polymerase.

Figure 2 with 1 supplement see all
Download asset Open asset
DinB2 and DinB3 overexpression confers antibiotic resistance through a distinct mutagenic profile.

(A and B) Rifampicin resistance (rifR) frequency in indicated strains in absence (blue) or presence (red) of inducer (50 nM anhydrotetracycline [ATc]). Results shown are means (± SEM) of data obtained from biological replicates symbolized by gray dots. Stars above bars mark a statistical difference with the reference (same strain without inducer) (***, p<0.001). Pie charts and bar chart in (A) shows the relative and absolute frequencies of nucleotide changes, represented with colors, detected in rpoB of indicated strains rifR in presence of inducer (50 nM ATc). The number of sequenced rifR is given in the center of each pie chart. (C) Location and relative frequency in % of mutated nucleotides in rpoB found in empty (blue), tet-dinB2 (red), or tet-dinB3 (orange) rifR. (D) Absolute frequency of the main rpoB mutations found in indicated strains in presence of 50 nM ATc. Empty = empty vector, tet = ATc-inducible promoter, DinB2=M. smegmatis DinB2, DinB3=M. smegmatis DinB3, ST = streptavidin tag, D107A=catalytically inactive M. smegmatis DinB2.

Rifampicin resistance is conferred by substitution mutations in the rifampin resistance determining region (RRDR) of the rpoB gene. To define the mutation spectrum stimulated by DinB2 and DinB3, we sequenced the RRDR of rifR colonies selected after overexpression of these polymerases (Figure 2A). In the strain carrying the empty vector, 38% of mutations conferring rifR were G>A or C>T and 24% were A>G or T>C, with a minority of other mutations, consistent with our prior report (Dupuy et al., 2022). DinB3 overexpression enhanced the relative frequency (expressed as %) of A>G or T>C mutations by 2.7-fold and the absolute frequency (expressed as mutants/108 CFU) by 11-fold with a minimal effect on other mutation types (Figure 2A). In contrast, DinB2 overexpression elicited a more diverse spectrum of mutations. Overexpressed DinB2 enhanced the relative frequency of A>C or T>G and G>T or C>A mutations by 2.8- and 2.3-fold, respectively. In addition, a previously undetected mutation type emerged: a trinucleotide mutation (G TC >T GT) spanning two codons and detected in 6.8% of sequenced rifR colonies. Overall, the absolute frequency of all mutation types was increased at least threefold after DinB2 overexpression, revealing the ability of the polymerase to stimulate diverse mutation types, in contrast to DinB1 (Dupuy et al., 2022).

Mapping of the DinB2 or DinB3 stimulated rpoB mutations onto the RRDR sequence revealed that 57% of the mutations incorporated by DinB3 were localized in the second nucleotide of the His442 codon vs 23% at this position in the control (Figure 2C). The predominant missense change at this codon was CAC>CGC (His>Arg) (Figure 2D), the same mutation stimulated by DinB1 (Dupuy et al., 2022). The absolute frequency of this mutation was increased 12-fold after DinB3 overexpression (Figure 2D). In contrast, DinB2 associated mutations were more widely distributed in the RRDR, particularly at His442 (CAC>CGC, >CCC, or >TAC), Ser438 (TCG>TTG or >TGG), Leu437/Ser438 (CTG TCG>CTT GTG), Leu449 (CTG>CCG), Asn435 (AAC>AAA), and Asp432 (GAC>GGC) (Figure 2C and D).

Overall, our results show that DinB2 and DinB3 can mediate rifampicin resistance but with different mutagenic profiles, with DinB2 driving a broader mutation spectrum. In comparison to our prior results with DinB1 and DnaE2, these data indicate that the mutagenic activities of DinB3 and DinB1 are relatively narrow and focused on His442, whereas DinB2 is similar to DnaE2 in its wider mutagenic spectrum.

DinB2 is highly prone to backward slippage in runs of A and T in vitro

Our previous study revealed that mycobacterial DinB1 mediates –1 and +1 frameshift mutations in runs of homo-oligonucleotides in vivo through a slippage activity of the polymerase (Dupuy et al., 2022). To determine whether DinB2 might have similar properties, we measured the ability of DinB2 to perform slippage in vitro. We reacted purified recombinant DinB2 with a 5' 32P-labeled primer-template DNA in the presence of dTTP or dATP. The DNA substrate consisted of a 13 bp duplex and a 5'-template tail in which the length of the homo-oligonucleotide contains 4, 6, or 8A or T followed by 3C or 3G (Figure 3A and B).

Figure 3 with 1 supplement see all
Download asset Open asset
DinB2 efficiently promotes –1 and +1 frameshifts in short and long runs of A and T.

(A and B) Reaction mixtures containing 10 mM Tris-HCl, pH 7.5, 5 mM MnCl2,1 pmol 5' 32P-labeled primer-template DNAs with indicated runs in the template strand (depicted below, and included as indicated above the lanes), 125 µM dTTP and ddGTP as specified, and 10 pmol DinB2 were incubated at 37°C for 15 min. The reaction products were analyzed by urea-PAGE and visualized by autoradiography. DinB2 was omitted from reactions in lanes –. (C–H) kanR frequencies in the indicated strains carrying the indicated mutation reporters in presence of inducer (50 nM ATc). Results shown are means (± SEM) of data obtained from biological replicates symbolized by gray dots. Stars above the bars mark a statistical difference with the reference strain (empty) (*, p<0.05; ***, p<0.001). Relative (pie chart) and absolute (bar chart) frequencies of nucleotide changes detected in kan of kanR cells represented with colors: pink=–1 or +1 frameshift in the homo-oligonucleotide run, blue=-2 frameshift in the run, green=+1 frameshift localized outside of the run and gray=no detected mutation. The number of sequenced kanR colonies is given in the center of each pie chart. Empty=empty vector, tet = Atc-inducible promoter, DinB2=M. smegmatis DinB2, DinB3=M. smegmatis DinB3, ST=streptavidin tag, D107A=catalytically inactive M. smegmatis DinB2.

Figure 3—source data 1

Original autoradiograms (Figure 3A and B).

https://cdn.elifesciences.org/articles/83094/elife-83094-fig3-data1-v1.zip

In presence of the A4 template and dTTP only, the reaction generated predominant +5 and minority +6 products (Figure 3A) that could be the consequence of either: the fill-in of the A4 run followed by the misincorporation of one or two Ts opposite C; or a backward slippage reaction catalyzed by DinB2 incorporating one or two extra Ts in the 4A template tract. DinB2’s slippage activity was more evident during synthesis on the A6 primer-template, generating a ladder of products elongated by 8–17 nucleotides. Progression to the A8 template generated a longer array of slippage products extended by 11–23 nucleotides. Similar results were obtained with template runs of 6 or 8 Ts (Figure 3B).

The finding that DinB2 is capable of iterative slippage synthesis on a homo-oligomeric tract when the only dNTP available is that templated by the homo-oligomer does not reflect the situation in vivo where DinB2 will have access to the next correctly templated dNTP. To query whether provision of the next templated nucleotide in vitro suppresses slippage, we included a dideoxy NTP (ddNTP): either ddGTP templated by the run of three C nucleotides following the A4, A6, and A8 tracts or ddCTP templated by the run of three G nucleotides flanking the T4, T6, and T8 tracts. ddNTPs are employed to force termination upon incorporation of the first templated nucleotide following the homo-oligomeric tract.

Inclusion of a templated ddGTP with dTTP generated a single +5 extension product on the A4 template, as expected for error-free addition of four dT nucleotides and a single terminal ddG (Figure 3A). Similarly, DinB2 reaction with the T4 template in the presence of ddCTP and dATP yielded a predominant +5 extension product reflecting 4 cycles of templated dA incorporation and a single terminal ddC addition (Figure 3B). In both cases, the +6 extension products seen with dTTP or dATP only (putative slippage) were suppressed by inclusion of a templated ddNTP.

The evidence for slippage was fortified by the effects of ddGTP on DinB2 activity on the A6 and A8 templates, where ladders of extensions products longer than +7 or +9 nucleotides (the expected results of error-free templated synthesis) were evident (Figure 3A). Up to 10 extra dTMP additions were detected on the A8 template even when the next templated nucleotide was available. Similar results applied to the T8 template, whereby ddCTP shortened but did not eliminate the slippage ladder seen with dATP alone (Figure 3B). Up to eight extra dAMP additions were observed on the T8 template in the presence of ddCTP. These experiments reveal that DinB2 is prone to backward slippage in A and T runs, a property that is exacerbated in longer homo-oligonu cleotide tracts. The heterogeneous size distribution of the slippage ladder is consistent with either of two scenarios: (1) multiple slippage cycles in which the primer 3’-OH end realigns backward on the template by a single nucleotide; and (2) one or several cycles of backward realignment of the primer 3’-OH on the template by more than one nucleotide (the upper limit being the length of the template homo-oligomeric tract) followed by fill-in to the end of the homo-oligomeric tract. It is noteworthy that reaction of DinB2 with the T4, T6, and T8 templates in the presence of dATP and ddCTP also generated a minor elongation product that was 1-nucleotide shorter than the predominant error-free ddC-terminated species, suggestive of a single cycle of forward slippage on the T runs prior to terminal ddC incorporation.

DinB2 promotes –1 and +1 frameshifts in short and long runs of A and T in vivo

The slippage activity of DinB2 in homo-oligonucleotide runs detected in vitro suggests that the polymerase may incorporate FS mutations in vivo. To test this possibility, we used a reporter tool developed in our previous study (Dupuy et al., 2022): a chromosomal integrated kan gene conferring the resistance to kanamycin inactivated by an out-of-frame homo-oligonucleotide run immediately downstream of the start codon. FS in the run can restore a functional reading frame and the mutation frequency can be quantified on kanamycin agar. To determine the effect of sequence specificity as well as run size, we used various reporters carrying different runs: 4T (kan::4T), 4A (kan::4A), 5T (kan::5T), 5A (kan::5A), 7T (kan::7T), 7A (kan::7A), 8T (kan::8T) and 8A (kan::8A).

Using the 4T reporter, we detected 6.6 kanamycin-resistant colonies (kanR) per 108 CFU in the control strain, carrying the empty vector (Figure 3C). A majority of the sequenced kanR had a –1 FS mutation in the run of Ts. The overexpression of DinB2 increased the kanR frequency more than 100-fold, compared to the control strain, but DinB3 had no effect. All kanR obtained after DinB2 overexpression had –1 FS localized in the kan 4T run, revealing that the polymerase strongly promotes this kind of mutation. The increase of the mutation frequency was reduced by DinB2 active site mutation D107A (Figure 3D), indicating that DinB2 directly incorporates –1 FS through its polymerase activity. The –1 FS mutation frequency observed with the 7T reporter in the control strain was much higher than with the 4T reporter (6.6 vs 3796 kanR/108 CFU), with 100% of the sequenced kanR events containing a –1 FS in the run (Figure 3E), revealing that the size of the run strongly impacts the spontaneous FS mutation frequency in mycobacteria. Even with this level of background, overexpression of DinB2 enhanced the –1 FS mutation frequency in the 7T run by 19-fold. A similar pattern was observed with runs of 4A and 7A (Figure 3—figure supplement 1A and B).

With 5T and 5A reporters, designed to detect +1 FS, the strain carrying the empty vector had 7 and 24 kanR/108 CFU, respectively, with 45% (5T) and 90% (5A) of sequenced kanR isolates having +1 FS mutations in the run (Figure 3F and Figure 3—figure supplement 1C). Whereas DinB3 overexpression had no effect on the kanR frequency, DinB2 overexpression increased the +1 FS frequency by 100-fold in both runs (Figure 3F and Figure 3—figure supplement 1C) and this effect was dependent on its polymerase activity (Figure 3G). An increase of the run size (5 vs 8) enhanced the spontaneous +1 FS detected in the run of A by 1000-fold and these events were still induced by DinB2 overexpression (Figure 3—figure supplement 1D). In contrast, the kanR frequency in the strain carrying the empty vector and the 8T reporter was similar to the frequency observed with the 5T reporter, but in contrast to the 5T, the sequenced kanR had no kan mutations (Figure 3H), potentially because the encoded three amino acid insertion impairs the function of the Aph protein or expression of the kanR gene. Despite this finding, we nevertheless observed an eightfold increase of the kanR frequency after DinB2 overexpression which was due to an accumulation of –2 FS mutations in the 8T run, revealing that DinB2 can also stimulate –2 FS.

Together, these results reveal that the size of homo-oligonucleotide runs strongly impacts the frequency of spontaneous FS mutations incorporated in A and T runs of the mycobacterial chromosome and that DinB2 can catalyze +1, –1, and –2 FS mutagenesis in these low complexity regions.

DinB2 slippage activity is enhanced on C and G homo-oligonucleotide templates

We proceeded to assay DinB2 with a series of primer-templates containing runs of G or C in vitro. Similar to our findings with 4A and 4T runs, DinB2 synthesis over the 4C run generated +5 and+6 products (Figure 4A). However, the polymerase was more prone to slip on the 4G template, generating a cluster of labeled primers extended by 6–9 cycles of dCMP addition (Figure 4B). Slippage on G4 was suppressed completely by inclusion of ddATP, which converted the ladder seen with dCTP alone into a single +5 extension product (Figure 4B). With the C4 template, inclusion of ddTTP altered the electrophoretic mobility of the error-free ddT-terminated +5 extension product vis-à-vis the +5 dG extension product arising via a single cycle of slippage (Figure 4A). A minor +6 extension product in the presence of ddTTP is evidence of residual +1 slippage on the C4 run in the presence of the next templated nucleotide.

Figure 4 with 2 supplements see all
Download asset Open asset
DinB2 slippage activity is enhanced on C and G homo-oligonucleotide templates.

(A and B) Reaction mixtures containing 10 mM Tris-HCl, pH 7.5, 5 mM MnCl2,1 pmol 5' 32P-labeled primer-template DNAs with indicated runs in the template strand (depicted below, and included as indicated above the lanes), 125 µM dGTP and ddTTP or dCTP and ddATP as specified, and 10 pmol DinB2 were incubated at 37°C for 15 min. DinB2 was omitted from reactions in lanes –. The reaction products were analyzed by urea-PAGE and visualized by autoradiography. The positions of the 13-mer primer strand and 5' 32P-labeled 40-mer and 50-mer oligonucleotide size markers analyzed in parallel are indicated on the right. (C–D) kanR frequencies in the indicated strains carrying indicated mutation reporters in presence of inducer (50 nM anhydrotetracyclin [ATc]). Results shown are means (± SEM) of data obtained from biological replicates symbolized by gray dots. Stars above the means mark a statistical difference with the reference strain (empty) (***, p<0.001). Relative (pie chart) and absolute (bar chart) frequencies of nucleotide changes detected in kan of kanR cells represented with colors: pink=–1 or +1 frameshift (FS) in the homo-oligonucleotide run, green=–1 or +1 FS localized outside of the run, light blue=+2 FS localized outside of the run, dark blue=>+2 insertion, brown=bases substitution mutation and gray=no detected mutation. The number of sequenced kanR colonies is given in the center of each pie chart. Empty=empty vector, tet=Atc-inducible promoter, DinB2=M. smegmatis DinB2, DinB3=M. smegmatis DinB3.

Figure 4—source data 1

Original autoradiograms (Figure 4A and B).

https://cdn.elifesciences.org/articles/83094/elife-83094-fig4-data1-v1.zip

Increasing the template G-run or C-run to 6 or 8 nucleotides strongly enhanced the slippage activity of DinB2, generating +12 to >50 products. Although the very longest slippage products were suppressed or diminished by inclusion of ddATP or ddTTP, DinB2 continued to synthesize long products by iterative slippage that contained tracts of 12 to ~30 dCMPs or dGMPs (Figure 4A and B).

The finding that DinB2 is more slippery during DNA synthesis across a template G run than on A runs or T runs of equivalent length vitiates the simple hypothesis that recession and realignment of the primer terminus on the template homo-oligonucleotide run is dictated by the thermodynamics of base pairing. If this were the case, we would expect the more weakly paired primer A-tract/template T-tract and primer T-tract/template A-tract configurations in Figure 3 to be more slippery than the more stably base-paired primer C-tract/template G-tract situation in Figure 4. That the opposite scenario applies suggests that DinB2 itself is uniquely geared to slip on G- or C-runs.

These results show that DinB2 is more prone to slippage in vitro than DinB1, that was assessed using a similar assay in our previous study (Dupuy et al., 2022). We also found that the polymerase domain of M. smegmatis Pol1 (Ghosh et al., 2020), homologous to the Klenow polymerase domain of E. coli Pol1, was weakly prone to slippage compared to DinB2 (Figure 4—figure supplement 1).

DinB2 slippage in G or C runs is size restricted in vivo

By using the in vivo kan reporter, we investigated the ability of DinB2 and DinB3 to incorporate FS in runs of G. The control strain (empty vector) carrying the 4G reporter had 21.7 kanR per 108 CFU and 35% of sequenced kanR isolates had a –1 FS in the 4G run (Figure 4C). The control strain which carries the 5G reporter showed 79.5 kanR per 108 CFU and +1 FS in the run were found in 67% of sequenced kanR isolates (Figure 4D). Overexpression of dinB3 did not significantly affect –1 FS or +1 FS mutations localized to the 4G or 5G runs. However, dinB2 overexpression enhanced –1 FS frequency in the 4G run by 100-fold and a non-statistically significant increase of +1 FS frequency was observed in the 5G run.

Results in Figure 4A and B show that DinB2 mostly catalyzes the synthesis of long slippage products, up to 60 nucleotides. These longer products, if they exist in vivo, would not be detected by our kan reporter because they will inactivate the Aph enzyme produced by the kan resistance cassette. To investigate if long slippage products are incorporated by DinB2 in vivo, we integrated a sacB gene carrying a run of 6C (sacB::6C) or 9C (sacB::9C) upstream of its start codon into the M. smegmatis genome. A functional sacB is lethal in presence of sucrose. Selection for sucrose resistance (sucR) would reveal all mutagenic events inactivating sacB, including putative long slippage events. In the strain carrying the empty vector and the sacB::6C reporter, sucR colonies were detected at a frequency of 1/10,000 (Figure 4—figure supplement 2A). The majority of the sucR clones did not have mutation in the 300 first bp of sacB, but 15% had a +1 FS in the 6C run and 15% had a substitution generating a stop codon. DinB2 overexpression increased sucR frequency by 10-fold and half of the sequenced colonies had +1 FS localized in the 6C run, 7% had a –1 FS in the run, and 7% a substitution confirming that DinB2 promotes both –1 and +1 FS in homo-oligonucleotide runs. However, the incorporation of long slippage products was not detected. In the control strain carrying the sacB::9C reporter, 3% of bacteria were sucR. Among them, 78% had a –1 FS in the run and 22% had a +1 FS in the run (Figure 4—figure supplement 2B). Overexpression of DinB2 did not increase the sucR frequency, which could be due to the high level of background, but the proportion of +1 FS in the run detected among sucR was enhanced by 2.5-fold. These results indicate that DinB2 does not incorporate long slippage products in vivo in these experimental conditions. We cannot exclude the possibility that the synthesis of long slippage tracks by DinB2, leading to large DNA loops, is lethal for the cell. These results also reveal the very high spontaneous FS frequency (>1%) in runs of nine homo-oligonucleotides.

DinB2 is less prone to slippage in RNA polymerase mode

DinB2 is the founder of a clade of Y-family DNA polymerase that is naturally adept at incorporating ribonucleotides, by virtue of a leucine in lieu of a canonical aromatic steric gate (Ordonez et al., 2014). Incorporation of rNTPs in DNA can lead to genome instability, replication fork blockage, and an increase of the mutation rate including FS (Kellner and Luke, 2020). To investigate if the DinB2 phenotypes reported in this work are due to its ribonucleotide utilizing activity, we constructed a restrictive steric gate mutant of DinB2 (L14F) which restores dNTP selectivity (Ordonez et al., 2014). The two proteins were expressed at the same level after inducer addition (Figure 5A). Overexpression of DinB2L14F resulted in a more substantial growth delay (Figure 5B) and enhanced cell death (Figure 5C) than overexpression of the WT protein. Induction of the DNA damage response (Figure 5D), substitution mutations (Figure 5E), or FS mutations (Figure 5F and G) were similar after overexpression of DinB2 or DinB2L14F. These results show that the mutagenesis phenotypes that accompany DinB2 overexpression are not due to ribonucleotide utilization by DinB2. These data also suggest that FS are directly incorporated by DinB2 and are not the secondary consequence of ribonucleotide excision repair (Schroeder et al., 2017).

Figure 5
Download asset Open asset
DinB2 does not slip in RNA polymerase mode.

(A) Anti-streptavidin/RpoB immunoblots from indicated strains after 16 hr of inducer treatment (50 nM anhydrotetracycline [ATc]). Average and SEM of RpoB normalized band intensities (n=3) are given below the image of a representative blot. (B) Growth of indicated strains in presence of 50 nM ATc. (C) Viability of indicated strains after 24 hr of inducer treatment (50 nM ATc). (D) Anti-RecA/RpoB immunoblots from indicated strains after 24 hr of inducer treatment (50 nM ATc). Average of normalized band intensities, expressed relative to the empty condition, is given below the image of a representative blot. (E) Rifampicin resistance (rifR) frequency in indicated strains in absence (blue) or presence (pink) of 50 nM ATc. (F and G) kanR frequencies in the indicated strains carrying indicated mutation reporters in presence of 50 nM ATc. Results shown are means (± SEM) of data obtained from biological replicates symbolized by gray dots or biological triplicates for (B). Stars above the means mark a statistical difference with the reference strain (B, C, F, and G: empty or E: same strain without inducer). Lines connecting two strains show a statistical difference between them. (**, p<0.01; ***, p<0.001). (H and I) Reaction mixtures containing 10 mM Tris-HCl, pH 7.5, 5 mM MnCl2, 1 pmol 5' 32P-labeled primer-template DNAs with indicated runs in the template strand (depicted below, and included as indicated above the lanes), 125 µM rUTP and ddGTP or rCTP and ddATP as specified, and 10 pmol DinB2 were incubated at 37°C for 15 min. DinB2 was omitted from reactions in lanes –. The reaction products were analyzed by urea-PAGE and visualized by autoradiography. The positions of the 13-mer primer strand and 5' 32P-labeled 40-mer and 50-mer oligonucleotide size markers analyzed in parallel are indicated on the right. Empty=empty vector, tet=Atc-inducible promoter, DinB2=M. smegmatis DinB2, ST=streptavidin tag, L14F=steric gate mutant of M. smegmatis DinB2.

Figure 5—source data 1

Uncropped immunoblots (Figure 5A and D) and original autoradiograms (Figure 5H and I).

https://cdn.elifesciences.org/articles/83094/elife-83094-fig5-data1-v1.zip

DinB2 can incorporate at least 16 consecutive ribonucleotides during primer extension on a DNA template (Ordonez et al., 2014). We queried whether DinB2 was prone to slippage on homo-oligonucleotide run templates when acting in RNA polymerase mode. DinB2 catalyzed mostly 4, 6, or 8 cycles of rUMP addition on the A4, A6, or A8 template (Figure 5H). The reaction on the A4 run revealed a very little extension to +5 that contrasts with the predominance of the +5 addition product generated by DinB2 in DNA polymerase mode (Figure 3A). Moreover, unlike the analogous DNA polymerase reaction with dTTP (Figure 3A), in the presence of rUMP we detected no slippage products longer than +8 and +10 on the A6 or A8 templates, respectively (Figure 5H). When ddGTP was added along with rUTP, DinB2 synthesized the expected +5, +7, and +9 ddG-terminated species (Figure 5H). Similar results were obtained when DinB2 was reacted with the G4, G6, and G8 primer-templates in the presence of rCTP (Figure 5I). This experiment reveals that DinB2 is much less slippery when utilizing rNTPs.

Metal-dependent switch in DinB2 slippage

Previous studies showed that DinB2 catalyzes a broader spectrum of deoxynucleotide misincorporations with manganese than magnesium (Ordonez and Shuman, 2014). Metal mixing experiments revealed that low ratios of manganese to magnesium sufficed to switch DinB2 to its more mutagenic mode (Ordonez and Shuman, 2014). This raises the question of how DinB2 behaves with respect to slippage in presence of manganese, magnesium, or both divalent cations. The in vitro experiments presented above were conducted in presence of 5 mM manganese only (Figures 3A, B—5H and I). The experiment in Figure 6A using the A6 primer-template entailed mixing 5 mM magnesium with increasing concentrations of manganese. We observed a manganese binary switch in the product distribution, from a no-slip state with magnesium alone (where the major product is a +7 mis-addition) to a slipped state in which DinB2 catalyzed 8 or more cycles of dTMP addition (Figure 6A). This transition was evident at 1 mM manganese, that is, a 1:5 ratio of Mn2+ to Mg2+. The length of the slippage tract, which was 8–10 nucleotides at 1 mM manganese, increased to 8–14 nucleotides at 2 mM manganese, and saturated at 8–17 nucleotides at 3–5 mM manganese. Note that mixing manganese and magnesium did not diminish the residual slippage events on the A6 template in the presence of ddGTP (Figure 6A). A similar metal-dependent switch in DinB2 slippage behavior was seen in experiments using the G6 template (Figure 6B), where a finer titration highlighted a transition from no-slippage to slippage at 0.5 mM manganese (a 1:10 ratio of Mn2+ to Mg2+). Magnesium did not impact the relatively high level of slippage on the G6 template that was seen in the presence of ddATP. We surmise that DinB2 has a preference for manganese occupancy of at least one of its two metal-binding sites (Ordonez and Shuman, 2014) when both magnesium and manganese are present, and that this occupancy suffices to shift DinB2 to a slippage mode.

Figure 6
Download asset Open asset
Metal-dependent switch in DinB2 activities.

(A and B) Reaction mixtures containing 10 mM Tris-HCl, pH 7.5, 1 pmol 5' 32P-labeled primer-template DNAs with an A6 or G6 run in the template strand (depicted below), divalent cations and nucleotides (125 µM) as specified above the lanes, and 10 pmol DinB2 were incubated at 37°C for 15 min. DinB2 was omitted from reactions in lanes –. The reaction products were analyzed by urea-PAGE and visualized by autoradiography. (C) Anti-streptavidin/RpoB immunoblots from indicated strains cultivated with indicated concentrations of MnCl2 (in mg/L) after 16 hr of inducer treatment (50 nM anhydrotetracyclin [ATc]). Average and SEM of normalized band intensities (n=3) are given below the image of representative blot. (D and E) Viability of indicated strains after 24 hr of inducer treatment (50 nM ATc) in presence of indicated concentration of MnCl2 (in mg/L). (F and G) kanR frequencies in the indicated strains carrying indicated mutation reporters in presence of inducer (50 nM ATc) and indicated concentration of MnCl2 (in mg/L). Results shown are means (± SEM) of data obtained from biological replicates symbolized by gray dots. Relative frequencies of nucleotide changes detected in kan of kanR cells are represented with colors: pink=–1 or +1 frameshift in the homo-oligonucleotide run, brown = substitution mutations. The number of sequenced kanR colonies is given in the center of each pie chart. Stars above the means mark a statistical difference with the reference strain: same strain untreated with Mn in (D) and (E), empty with same Mn treatment in (F) and (G) (**, p<0.01; ***, p<0.001). Empty = empty vector, tet = Atc-inducible promoter, DinB2=M. smegmatis DinB2, ST = streptavidin tag, D107A=catalytically inactive M. smegmatis DinB2.

Figure 6—source data 1

Original autoradiograms (Figure 6A and B) and uncropped immunoblots (Figure 6C).

https://cdn.elifesciences.org/articles/83094/elife-83094-fig6-data1-v1.zip

Modulation of DinB2 activity in vivo by manganese

To investigate the consequence of the divalent cations nature on the DinB2 activity in vivo, we added increasing concentrations of manganese to Middlebrook 7H9 culture medium which contains 50 mg/L magnesium, with or without induced DinB2 overexpression. Addition of manganese did not impact the DinB2 expression level (Figure 6C). Whereas addition of manganese had no effect on the viability of the strain carrying the empty vector, it decreased the viability of the strain overexpressing DinB2 (Figure 6D), a phenotype that was reversed by inactivation of DinB2 polymerase activity (Figure 6E). We next measured the impact of manganese on DinB2 stimulated frameshifting in vivo. We confirmed that, in absence of manganese, overexpression of DinB2 increased –1 FS or +1 FS localized in runs of 4T or 5T by 100-fold, but addition of manganese did not alter frameshifting (Figure 6F and G). Manganese addition also did not impact the DinB2-dependent mutation frequency or type of mutation detected with the sacB reporter (Figure 4—figure supplement 2). These results indicate that manganese exacerbates the deleterious polymerase activity of DinB2 in vivo but that this effect is likely due to other Mn-dependent activities of DinB2.

Discussion

The mutagenic properties of DinB2

DinB2 can execute several types of mutagenesis in vitro, including ribonucleotide insertion in DNA, utilization of 8-oxoguanine, and nucleotide misincorporation (Ordonez and Shuman, 2014; Ordonez et al., 2014; Sharma and Nair, 2012). However, the role of DinB2 in mutagenesis in vivo was unknown and how any DinB2-dependent mutagenic pathway might overlap with the recently described DinB1 pathway had not been examined.

We found that forced expression of DinB3 causes a similar spectrum of base substitutions as DinB1 (Dupuy et al., 2022), mainly A>G or T>C mutations. These mutations are focused on single codon in RpoB to produce the His442>Arg mutation. However, DinB2 has a very different mutagenic spectrum, inducing a diversity of mutation types, particularly A>C or T>G and G>T or C>A. This DinB2 effect is dependent on its polymerase activity, whereas we did not express a DinB3 active site mutant. These mutations are the typical signature of 8-oxoguanine presence in DNA (Dupuy et al., 2020; van Loon et al., 2010) and are consistent with the 8-oxoguanine handling capabilities of DinB2 (Ordonez and Shuman, 2014). Moreover, we recently proposed that DinB2, together with other TLS polymerases, contributes to fluoroquinolone bactericidal action by incorporating 8-oxoguanine in M. smegmatis genome when the MutT system is inactivated (Dupuy et al., 2020). The mutagenesis data presented here further implicates DinB2 in utilizing 8-oxoguanine in vivo under conditions of endogenous or exogenous oxidative stress.

The four most frequent RpoB mutations induced by DinB2 are H442R, H442P, L449P, and S438L. Whereas H442R and S438L mutations are also induced by DinB1, DinB3, or DnaE2 (Dupuy et al., 2022), H442P (CAC>CCC) and L449P (CTG>CCG) mutations are more specific to DinB2. In Mtb, the equivalent RpoB mutations are H445P and L452P which comprise 0.1% and 1.2% of rifR mutations in clinical isolates, respectively (World Health Organisation, 2021). Taken together, the mutational spectrum data for DinB1, DnaE2, and DinB2 indicate a division of labor between TLS polymerases in the generation of rifampicin resistance in Mtb clinical strains. However, unlike deletion of dnaE2, loss of DinBs does not decrease rifR acquisition in M. smegmatis or Mtb in various tested conditions (Dupuy et al., 2022; Kana et al., 2010), either suggesting redundancy or that a relevant mutagenic stress is missing in culture.

DinB2 as a catalyst of frameshifting

Because of the high FS frequency observed in mycobacteria (Gupta and Alland, 2021; Springer et al., 2004) as well as the link of these mutations with antibiotic tolerance in Mtb (Bellerose et al., 2019; Safi et al., 2019), it is crucial to identify the mechanisms of frameshifting in mycobacteria. Our previous study revealed that DinB1 and DnaE2 promote FS in homo-oligonucleotide runs (Dupuy et al., 2022). Here, we showed that DinB2, but not DinB3, also promotes –1 and +1 FS mutagenesis in short (4–5) and long (6–8) homo-oligonucleotide runs in vivo when overexpressed. Compared to DinB1 and DnaE2, overexpression of dinB2 more efficiently promotes FS mutations in short runs. DinB2 is also more prone to slippage in vitro compared to DinB1 (Dupuy et al., 2022) or Pol1 (Figure 4—figure supplement 1), generating long backward slippage products (+5 to ~+60) in homo-oligomeric template runs of 6 or 8 nucleotides (Figures 3A, B, 4A and B).

By using dinB1 and dnaE2 mutants, we previously showed that DinB1 mediates spontaneous –1 FS mutagenesis and DnaE2 is the primary agent of DNA damage-induced –1 FS mutations in M. smegmatis (Dupuy et al., 2022). However, in that study dinB2 deletion, alone or in combination with the deletion of other TLS polymerases, did not impact FS mutagenesis in short homonucleotide runs (Dupuy et al., 2022). The present work suggests that environmental conditions, including dNTP/rNTP balance or metal ion availability, can influence the mutagenic activity of DinB2 (Figures 5 and 6). In addition, recent studies revealed that the dinB2 expression is under the control of at least two TetR family proteins (Msmeg_6564 and Msmeg_2295), and their deletion and/or overexpression induce dinB2 expression (Yang et al., 2012; Patra et al., 2021). Menaquinone, a respiratory lipoquinone, induces dinB2 expression more than 10-fold, by both inhibiting the Msmeg_2295 repression and acting as a direct inducer (Barman et al., 2022). Other experimental conditions will need to be tested to assess the impact of dinB2 deletion on mutagenesis in mycobacteria, including during infection or after menaquinone treatment, to confirm the phenotypes demonstrated here which are derived from induced expression of DinB2.

Frameshifts and homo-oligonucleotide runs in mycobacterial genomes

Our study is the first to systematically examine the influence of run length and nucleotide composition on frameshift mutations in mycobacterial homopolymeric sequences. Our data demonstrate that homo-oligonucleotide runs promote –1 and +1 FS mutations and that the length of the run strongly impacts the frequency of FS. Depending on the reporter, we obtained spontaneous FS frequency between 1/107 and 1/106 per CFU in short runs (4–5 nucleotides) and between 1/105 and 1/102 per CFU in long runs (7–9 nucleotides). Our study complements two in silico analyses showing that frameshifts, but not general mutation rates, are higher in mycobacterial genomes than other organisms (Springer et al., 2004) and that indels are significantly enriched in homo-oligonucleotide runs of clinical Mtb isolates (Gupta and Alland, 2021). A reasonable hypothesis explaining this remarkable frameshift frequency is the absence of a conventional MutS/L mismatch repair in mycobacteria. A study in yeast showed that FS increases 400-fold in a run of 14A relative to a run of 4A in a WT strain which contrasts with the 51,000-fold increase obtained in a mismatch repair mutant (msh2), confirming that mismatch repair is a key mechanism for correction of FS in homo-oligonucleotide runs, particularly in long tracks (Tran et al., 1997). However, in mycobacteria MutS/L is replaced by an alternative system which depends on NucS and corrects substitution mutations but not FS (Castañeda-García et al., 2017; Castañeda-García et al., 2020).

Recent data has highlighted the frequency of homopolymeric sequences in mycobacterial genomes and hinted at the functional consequences of sequence variation at these sites. Transient Mtb drug tolerance can be conferred by an FS mutation in a run of C (between 7C and 10C depending on the Mtb strain) of the glpK gene, encoding glycerol kinase (Bellerose et al., 2019; Safi et al., 2019). In addition, recent in silico analyses of genomes of Mtb clinical isolates identified frequent indels in homopolymeric tracts, some of which may have important virulence functions, revealing that reversible gene silencing mediated by FS is not restricted to the glpK-dependent antibiotic tolerance pathway (Gupta and Alland, 2021; Vargas et al., 2022). Based on the level of FS frequency we detected in homo-oligonucleotide runs, we believe that reversible gene silencing through frameshifting in long runs is a prevalent phenomenon in mycobacteria and that DinB1 and DinB2 could contribute to these indels. Further studies are needed to explore the biological functions of reversible gene silencing in mycobacteria, and the roles of DinBs in this process, particularly in genes that contain long homo-oligonucleotide runs.

A complex network of TLS polymerases in mycobacteria

Mycobacterial DinBs belong to three phylogenetic subfamilies (DinB1, DinB2, and DinB3), distinct from the E. coli DinB clade (Timinskas and Venclovas, 2019). Timinskas and Venclovas detected 144 bacterial species, almost all actinobacteria, encoding at least one DinB belonging to DinB1, DinB2, or DinB3 family. Among them, 78% encode DinB1, 33% DinB2 and 40% DinB3. DinB1, but not DinB2 or DinB3, interacts with the β clamp and confers tolerance to DNA damage (Dupuy et al., 2022; Kana et al., 2010), suggesting that DinB1 is the functional ortholog of E. coli DinB. Two models are proposed for TLS (Joseph and Badrinarayanan, 2020): (1) TLS at the fork during which TLS polymerase switches with the replicative polymerase and assists the replicative machinery in lesion bypass. (2) TLS behind the fork during which the replicative machinery skips past the lesion and continues synthesis downstream, the gap being subsequently filled by the TLS polymerase. E. coli DinB seems to mediate both of these pathways (Joseph and Badrinarayanan, 2020). A unifying model based on our data is the existence of a division of labor between mycobacterial DinBs, DinB1 mediating TLS at the fork, and DinB2/DinB3 involving in TLS behind the fork. DinB2 and DinB3 could also have other functions, not related to replication, including gap filling during DNA repair. It is intriguing that lower gene dosage of DinB2 in M. smegmatis confers susceptibility to dTTP-limiting conditions imposed by expression of a phage ribonucleotide reductase (Ghosh et al., 2015). This finding, together with the ability of DinB2 to utilize rNTPs (Ordonez et al., 2014), suggests that DinB2 could play an important role in quiescent cells, when dNTPs are limiting.

The role of DinBs in Mtb during host infection is still unknown. Deletion of dinB1 and dinB2 does not reduce bacterial survival in macrophages or in mice (Kana et al., 2010). However, redundancy with dnaE2, which is important for Mtb virulence (Boshoff et al., 2003), has been demonstrated for DNA damage tolerance (Dupuy et al., 2022) but never explored in the context of infection. Finally, the ability of mycobacterial DinBs, and particularly DinB2, to incorporate FS mutation in homo-oligonucleotide runs, suggests a role for these DNA polymerases in genome evolution through reversible FS in low complexity sequences (Bellerose et al., 2019; Safi et al., 2019; Safi et al., 2020).

Methods

Bacterial strains

Bacterial strains used in this work are listed in key resources table. E. coli strains were grown in Luria-Bertani medium at 37°C. M. smegmatis strains were cultivated at 37°C in Middlebrook 7H9 or 7H10 media (Difco) supplemented with 0.5% glycerol, 0.5% dextrose, and 0.1% Tween 80 (for 7H9 only). Hygromycin was used at 50 μg/mL.

Plasmids

Plasmids and primers used in this work are listed in key resources table. kan plasmids (–1 or +1 frameshift mutation reporters) were constructed as reported in Dupuy et al., 2022, using primers listed in key resources table. For sacB plasmids (unbiased frameshift mutation reporters), sacB was amplified using pAJF067 as a template and primers listed in key resources table and were cloned into pDB60 digested with EcoR1 using recombination-based cloning (In-Fusion, Takara). For dinB2 and dinB3 overexpression plasmids constructs, ORFs were amplified using primers listed in key resources table and M. smegmatis mc2155 genomic DNA as a template and were cloned into pmsg419 digested with ClaI using In-fusion cloning. Cloning was performed in E. coli DH5α and plasmids were introduced in M. smegmatis by electroporation.

Growth and cell viability

Bacteria were grown in absence of inducer (ATc) and diluted to a calculated OD600=0.001 in fresh medium supplemented with 50 nM ATc (indicated on the figure when other concentrations were used). OD600 was measured for 48 hr. Cultures were diluted in fresh medium supplemented with ATc to a calculated OD600=0.001 when OD600 reached a value around 1 and measured OD600 was corrected by the dilution factor. For viability experiments, ATc was added to exponential growth phase cultures for variable periods of time. Then, OD600 was measured and cells were washed in –ATc fresh medium and plated on –ATc agar medium for CFU enumeration. Viability was expressed in CFU number per optical density unit (CFU in 1 mL of a culture at OD600=1). For the experiments in Figure 6, MnCl2 was added to the +ATc culture, but not to the agar media. For agar medium growth experiments, log-phase cells were cultivated in liquid medium without inducer and serial dilutions were spotted (5 µL) on agar medium supplemented with 0, 2.5, 5, or 50 nM ATc. Plates were pictured after 72 hr of incubation at 37°C.

Western blot

Two mL aliquots of a log-phase culture at OD600 of 0.4, treated or not with ATc, were used for cell lysate preparation. Lysate preparation and protein separation was conducted as described in Dupuy et al., 2022. Blots were blocked and probed in 5% Omniblot milk (RecA detection) or 3% BSA (ST DinBs detection) in PBST (PBS buffer supplemented with 0.1% Tween 20). Proteins on blots were detected using anti-RpoB (Biolegend; 663905; AB_2566583), anti-RecA (Pocono Rabbit Farm & Laboratory, Wipperman et al., 2018), or anti-streptavidin (STII GenScript rabbit anti-NWSHPQFEK polyclonal antibody, Adefisayo et al., 2021) antibodies incubated at a 1:10,000 dilution for 1 hr and secondary horseradish peroxidase antibodies. Membranes were treated with Amersham ECL western blotting detection reagents (GE Healthcare) according to the manufacturer’s instructions Blots were imaged with an iBright FL1000 (Invitrogen). Band intensities were within the linear range of detection and were quantitated with ImageJ in relation to RpoB loading controls.

Substitution and frameshift mutation frequency

Log-phase bacteria cultured from a single colony were diluted to OD600 of 0.004 in fresh medium supplemented or not with 50 nM ATc and/or variable concentration of MnCl2. After 16 hr of culture (OD600 ~0.5), cells were concentrated 20-fold by centrifugation and pellet resuspension and 100 µL of a 10–6 dilution was plated on 7H10 agar whereas 200 µL was plated on 7H10 supplemented with 100 µg/mL rif (substitution mutations), 4 µg/mL Kan (–1 or +1 FS mutations), or 5% sucrose (unbiased FS mutations). Mutation frequency was expressed by the average number of selected colonies per 108 CFU from independent cultures. The number of independent cultures used to measure the mutation frequency of each strain is indicated by the number of gray dots in each graph. The mutation spectrum was determined by amplification and sequencing of the RRDR of the rpoB gene of isolated rifR colonies, the kan gene of isolated kanR colonies, or the sacB gene of isolated sucR colonies. Sequenced rifR, kanR, or sucR colonies were selected from at least three independent bacterial cultures (between 2 and 8 clones per independent culture were sequenced). The number of sequenced isolated colonies is given in the center of each pie chart. Primers used for amplification and sequencing are listed in key resources table. Mutation spectrum is expressed as relative frequency (percent of mutation types) or absolute frequency (number of each mutation type per 108 CFU obtained by multiplying the rifR or kanR frequency by the relative frequency of the mutation type).

In vitro DNA slippage assay

DinB2 and the C-terminal POL domain of Pol1 were produced in E. coli and purified as described previously (Ordonez et al., 2014; Ghosh et al., 2020). Protein concentrations were determined by using the Bio-Rad dye reagent with bovine serum albumin as the standard. A 5' 32P-labeled primer DNA strand was prepared by reaction of a 13-mer oligonucleotide with T4 polynucleotide kinase and [γ32P]ATP. The labeled DNA was separated from free ATP by electrophoresis through a nondenaturing 18% polyacrylamide gel and then eluted from an excised gel slice. The primer-templates for assays of DNA polymerase were formed by annealing the 5' 32P-labeled 13-mer pDNA strand to an unlabeled template strand at 1:3 molar ratio. The DNA mixtures were incubated serially at 65°C, 37°C, and 22°C to promote strand annealing. Polymerase reaction mixtures (10 µL) containing 50 mM Tris-HCl, pH 7.5, MgCl2 or MnCl2 as specified, 0.125 mM dNTP, ddNTP, or rNTP as specified, 1 pmol (0.1 µM) 32P-labeled primer-template DNA, and 10 pmol (1 µM) DinB2 were incubated at 37°C for 15 min. The reactions were quenched by adding 10 µL of 90% formamide, 50 mM EDTA, 0.01% bromophenol blue-xylene cyanol. The samples were heated at 95°C for 5 min and then analyzed by electrophoresis through a 40 cm 18% polyacrylamide gel containing 7.5 M urea in 44.5 mM Tris-borate, pH 8.3, 1 mM EDTA. The products were visualized by autoradiography.

Statistical analysis

Statistical tests were performed using Prism9 software (GraphPad) on ln-transformed data. Statistical differences in growth and viability experiments were tested using a two-way analysis of variance (ANOVA) and a Bonferroni post-test. Statistical differences in mutagenesis experiments were tested using a t-test to compare the means of two population groups or a one-way ANOVA and a Bonferroni post-test to compare the means of more than two population groups.

Appendix 1

Appendix 1—key resources table
Reagent type (species) or resourceDesignationSource or referenceIdentifiersAdditional information
Strain, strain background (Escherichia coli)DH5αThe Glickman lab
Strain, strain background (Mycobacterium smegmatis)Wild type (mc2155)Snapper et al., 1990PDS1
Strain, strain background (Mycobacterium smegmatis)∆recADupuy et al., 2020PDS353
Strain, strain background (Mycobacterium smegmatis)∆dnaE2Dupuy et al., 2020PDS139
Strain, strain background (Mycobacterium smegmatis)pmsg419This workMgm4062Available from Glickman Lab
Strain, strain background (Mycobacterium smegmatis)pRGM47This workmgm4063Available from Glickman Lab
Strain, strain background (Mycobacterium smegmatis)pRGM48This workmgm4072Available from Glickman Lab
Strain, strain background (Mycobacterium smegmatis)pRGM49This workmgm4073Available from Glickman Lab
Strain, strain background (Mycobacterium smegmatis)pRGM50This workmgm4074Available from Glickman Lab
Strain, strain background (Mycobacterium smegmatis)pDP69This workPDS416Available from Glickman Lab
Strain, strain background (Mycobacterium smegmatis)pDP70This workPDS417Available from Glickman Lab
Peptide, recombinant proteinDinB1Ordonez et al., 2014
Peptide, recombinant proteinDinB2Ordonez et al., 2014
Recombinant DNA reagentATc-on system vector (hygR, OriMyc)Lab Stockpmsg419Available from Glickman Lab
Recombinant DNA reagentMycob. integr. vector (StrepR, attP(L5))Lab StockpDB60Available from Glickman Lab
Recombinant DNA reagentpmsg419-dinB2 Strep tagThis workpRGM47Cloning primers: dinB2fw-dinB2rev1 (cloning enzyme site: ClaI);
available from Glickman Lab
Recombinant DNA reagentpmsg419-dinB2This workpRGM48Cloning primers: dinB2fw-dinB2rev2 (cloning enzyme site: ClaI);
dinB2=Msmeg_2294; available from Glickman Lab
Recombinant DNA reagentpmsg419-dinB2D107A Strep tagThis workpRGM49Cloning primers: dinB2fw-dinB2catrev +dinB2catfw-dinB2rev1
(cloning enzyme site: ClaI); available from Glickman Lab
Recombinant DNA reagentpmsg419-dinB2L14F Strep tagThis workpRGM50Cloning primers: dinB2fw-dinB2stericrev +dinB2stericfw-dinB2rev1
(cloning enzyme site: ClaI); available from Glickman Lab
Recombinant DNA reagentpmsg419-dinB3This workpDP69Cloning primers: ODP197-ODP298 (cloning enzyme site: ClaI);
available from Glickman Lab
Recombinant DNA reagentpmsg419-dinB3 Strep tagThis workpDP70Cloning primers: ODP197-ODP299 (cloning enzyme site: ClaI);
available from Glickman Lab
Recombinant DNA reagentpDB60 derivative with kan::3TDupuy et al., 2022pDP120
Recombinant DNA reagentpDB60 derivative with kan::3CDupuy et al., 2022pDP121
Recombinant DNA reagentpDB60 derivative with kan::3GDupuy et al., 2022pDP122
Recombinant DNA reagentpDB60 derivative with kan::3ADupuy et al., 2022pDP123
Recombinant DNA reagentpDB60 derivative with kan::4TDupuy et al., 2022pDP124
Recombinant DNA reagentpDB60 derivative with kan::4CDupuy et al., 2022pDP125
Recombinant DNA reagentpDB60 derivative with kan::4GDupuy et al., 2022pDP126
Recombinant DNA reagentpDB60 derivative with kan::4ADupuy et al., 2022pDP127
Recombinant DNA reagentpDB60 derivative with kan::6TDupuy et al., 2022pDP128
Recombinant DNA reagentpDB60 derivative with kan::6CDupuy et al., 2022pDP129
Recombinant DNA reagentpDB60 derivative with kan::6GDupuy et al., 2022pDP130
Recombinant DNA reagentpDB60 derivative with kan::6ADupuy et al., 2022pDP131
Recombinant DNA reagentpDB60 derivative with kan::7TThis workpDP132Cloning primers: ODP443-ODP445+ODP458-ODP444
(cloning enzyme site: EcoRI); available from Glickman Lab
Recombinant DNA reagentpDB60 derivative with kan::7CThis workpDP133Cloning primers: ODP443-ODP445+ODP459-ODP444
(cloning enzyme site: EcoRI); available from Glickman Lab
Recombinant DNA reagentpDB60 derivative with kan::7GThis workpDP134Cloning primers: ODP443-ODP445+ODP460-ODP444
(cloning enzyme site: EcoRI); available from Glickman Lab
Recombinant DNA reagentpDB60 derivative with kan::7AThis workpDP135Cloning primers: ODP443-ODP445+ODP461-ODP444
(cloning enzyme site: EcoRI); available from Glickman Lab
Recombinant DNA reagentpDB60 derivative with kan::9TThis workpDP136Cloning primers: ODP443-ODP445+ODP462-ODP444
(cloning enzyme site: EcoRI); available from Glickman Lab
Recombinant DNA reagentpDB60 derivative with kan::9CThis workpDP137Cloning primers: ODP443-ODP445+ODP463-ODP444
(cloning enzyme site: EcoRI); available from Glickman Lab
Recombinant DNA reagentpDB60 derivative with kan::9GThis workpDP138Cloning primers: ODP443-ODP445+ODP464-ODP444
(cloning enzyme site: EcoRI); available from Glickman Lab
Recombinant DNA reagentpDB60 derivative with kan::9AThis workpDP139Cloning primers: ODP443-ODP445+ODP465-ODP444
(cloning enzyme site: EcoRI); available from Glickman Lab
Recombinant DNA reagentpDB60 derivative with kan::5TDupuy et al., 2022pDP144
Recombinant DNA reagentpDB60 derivative with kan::5CDupuy et al., 2022pDP145
Recombinant DNA reagentpDB60 derivative with kan::5GDupuy et al., 2022pDP146
Recombinant DNA reagentpDB60 derivative with kan::5ADupuy et al., 2022pDP147
Recombinant DNA reagentpDB60 derivative with kan::8TThis workpDP148Cloning primers: ODP443-ODP445+ODP494-ODP444
(cloning enzyme site: EcoRI); available from Glickman Lab
Recombinant DNA reagentpDB60 derivative with kan::8CThis workpDP149Cloning primers: ODP443-ODP445+ODP495-ODP444
(cloning enzyme site: EcoRI); available from Glickman Lab
Recombinant DNA reagentpDB60 derivative with kan::8GThis workpDP150Cloning primers: ODP443-ODP445+ODP496-ODP444
(cloning enzyme site: EcoRI); available from Glickman Lab
Recombinant DNA reagentpDB60 derivative with kan::8AThis workpDP151Cloning primers: ODP443-ODP445+ODP497-ODP444
(cloning enzyme site: EcoRI); available from Glickman Lab
Recombinant DNA reagentpDB60 derivative with sacB::9CThis workpDP186Cloning primers: ODP593-ODP596+ODP597-ODP598
(cloning enzyme site: EcoRI); available from Glickman Lab
Recombinant DNA reagentpDB60 derivative with sacB::6CThis workpDP194Cloning primers: ODP593-ODP614+ODP615-ODP598
(cloning enzyme site: EcoRI); available from Glickman Lab
AntibodyAnti-RpoB (mouse monoclonal)Biolegend663905; AB_2566583(1:10,000 dilution)
AntibodyAnti-RecA (rabbit polyclonal)Wipperman et al., 2018Anti-RecA(1:10,000 dilution)
AntibodyAnti-streptavidin (rabbit polyclonal)GenScriptA00626STII GenScript rabbit anti-NWSHPQFEK; used at (1:10,000 dilution)
Sequence-based reagentfw dinB2IDTdinB2fwCAGAAAGGAGGCCATATGACCAAATGGGTGCTC
Sequence-based reagentrev dinB2+streptavidin tagIDTdinB2rev1AGGTCGACGGTATCGATACTACTTTTCGAACTGCG
GGTGGCTCCAGGTGCCTGCAGTGACAG
Sequence-based reagentrev dinB2IDTdinB2rev2AGGTCGACGGTATCGATGTGCTCGAGTTAGGTGCCTGCAGTGAC
Sequence-based reagentrev internal dinB2Msm with pol. dead mut. (D107A)IDTdinB2catrevGCCCAGATACGCCTCGGCCCAGCCCCACACCTCCAAC
Sequence-based reagentfw internal dinB2Msm with pol. dead mut. (D107A)IDTdinB2catfwGCCGAGGCGTATCTGGGC
Sequence-based reagentrev internal dinB2Msm with steric gate mut. (L14F)IDTdinB2stericrevGCAACTCCACCGAAGCAAAGAACTGGTCCAGATCGAC
Sequence-based reagentfw internal dinB2Msm with steric gate mut. (L14F)IDTdinB2stericfwTTTGCTTCGGTGGAGTTGC
Sequence-based reagentfw dinB3IDTODP297CAGAAAGGAGGCCATATGTTCGTGTCCGCTGC
Sequence-based reagentrev dinB3IDTODP298AGGTCGACGGTATCGCTAGTCCGGCAGCATGG
Sequence-based reagentrev dinB3+streptavidin tagIDTODP299AGGTCGACGGTATCGCTACTTTTCGAACTGCGGGT
GGCTCCAGTCCGGCAGCATGGG
Sequence-based reagentfw kanIDTODP443TCCAGCTGCAGAATTTCCCAAGGACACTGAGTCC
Sequence-based reagentrev kanIDTODP444GATAAGCTTCGAATTTTGCTGACTCATACCAGGC
Sequence-based reagentInternal rev kanIDTODP445CATAACACCCCTTGTATTACTG
Sequence-based reagentInternal fw kan (7T addition)IDTODP458ACAAGGGGTGTTATGTTTTTTTAGCCATATTCAACGGGAAACG
Sequence-based reagentInternal fw kan (7C addition)IDTODP459ACAAGGGGTGTTATGCCCCCCCAGCCATATTCAACGGGAAACG
Sequence-based reagentInternal fw kan (7G addition)IDTODP460ACAAGGGGTGTTATGGGGGGGAAGCCATATTCAACGGGAAACG
Sequence-based reagentInternal fw kan (7A addition)IDTODP461ACAAGGGGTGTTATGGAAAAAAAGCCATATTCAACGGGAAACG
Sequence-based reagentInternal fw kan (9T addition)IDTODP462ACAAGGGGTGTTATGTTTTTTTTTAGCCATATTCAACGGGAAACG
Sequence-based reagentInternal fw kan (9C addition)IDTODP463ACAAGGGGTGTTATGCCCCCCCCCAGCCATATTCAACGGGAAACG
Sequence-based reagentInternal fw kan (9G addition)IDTODP464ACAAGGGGTGTTATGGGGGGGGGAAGCCATATTCAACGGGAAACG
Sequence-based reagentInternal fw kan (9A addition)IDTODP465ACAAGGGGTGTTATGGAAAAAAAAAGCCATATTCAACGGGAAACG
Sequence-based reagentInternal fw kan (8T addition)IDTODP494ACAAGGGGTGTTATGTTTTTTTTAGCCATATTCAACGGGAAACG
Sequence-based reagentInternal fw kan (8C addition)IDTODP495ACAAGGGGTGTTATGCCCCCCCCAGCCATATTCAACGGGAAACG
Sequence-based reagentInternal fw kan (8G addition)IDTODP496ACAAGGGGTGTTATGGGGGGGGAAGCCATATTCAACGGGAAACG
Sequence-based reagentInternal fw kan (8A addition)IDTODP497ACAAGGGGTGTTATGGAAAAAAAAGCCATATTCAACGGGAAACG
Sequence-based reagentfw sacBIDTODP593TCCAGCTGCAGAATTAACCCATCACATATACCTGCCG
Sequence-based reagentInternal rev sacB (9C addition)IDTODP596GTTGGGGGGGGGCATCGTTCATGTCTCCTTTTTTATG
Sequence-based reagentInternal fw sacB (9C addition)IDTODP597ATGCCCCCCCCCAACATCAAAAAGTTTGCAAAACAAG
Sequence-based reagentrev sacBIDTODP598GATAAGCTTCGAATTACTATCAATAAGTTGGAGTCATTACC
Sequence-based reagentInternal rev sacB (6C addition)IDTODP614GTTGGGGGGCATCGTTCATGTCTCCTTTTTTATG
Sequence-based reagentInternal fw sacB (6C addition)IDTODP615ATGCCCCCCAACATCAAAAAGTTTGCAAAACAAG
Sequence-based reagentfw PCR screening and seq pmsg419 cloningIDTODP236CTCCCTATCAGTGATAGATAGGCTCTGG
Sequence-based reagentrev PCR screening and seq pmsg419 cloningIDTODP237CATGACCAACTTCGATAACGTTCTCGG
Sequence-based reagentfw PCR screening and seq pDB60 cloningIDTODP474TGATTCTGTGGATAACCGTATTACCGCCTTTG
Sequence-based reagentrev PCR screening and seq pDB60 cloningIDTODP475AAGGCCCAGTCTTTCGACTGAGC
Sequence-based reagentfw rpoB PCRIDTODP378CAAGAAGCTGGGCCTGAACGC
Sequence-based reagentrev rpoB PCRIDTODP379GCGGTTGGCGTCGTCGTG
Sequence-based reagentrpoB seqIDTODP380GAGCGTGTCGTGCGTGAG
Sequence-based reagentfw kan or sacB PCRIDTODP476TGGCCTTTTGCTGGCCTTTTGC
Sequence-based reagentrev kan PCRIDTODP477TTCAACAAAGCCGCCGTCCC
Sequence-based reagentkan seqIDTODP479ACTGAATCCGGTGAGAATGG
Sequence-based reagentrev sacB PCRIDTODP172TTAGACGTAATGCCGTCAATCGTC
Sequence-based reagentsacB seqIDTODP474TGATTCTGTGGATAACCGTATTACCGCCTTTG
Sequence-based reagent5' 32P-labeled primer DNA strandIDTSG-FS1CGTGTCGCCCTTC
Sequence-based reagentUnlabeled template strand (4T)IDTSG-FS1GGGTTTTGAAGGGCGACACG
Sequence-based reagentUnlabeled template strand (6T)IDTSG-FS1GGGTTTTTTGAAGGGCGACACG
Sequence-based reagentUnlabeled template strand (8T)IDTSG-FS1GGGTTTTTTTTGAAGGGCGACACG
Sequence-based reagentUnlabeled template strand (4A)IDTSG-FS1CCCAAAAGAAGGGCGACAC
Sequence-based reagentUnlabeled template strand (6A)IDTSG-FS1CCCAAAAAAGAAGGGCGACAC
Sequence-based reagentUnlabeled template strand (8A)IDTSG-FS1CCCAAAAAAAAGAAGGGCGACAC

Data availability

Further information and requests for resources and reagents should be directed to and will be fulfilled by Dr. Michael Glickman (Glickmam@mskcc.org). Plasmids and strains generated in this study will be made available on request. All data generated in this study are presented in the figures and tables and the underlying data supporting all figures is provided as Source data 1.

References

Decision letter

  1. Bavesh D Kana
    Senior and Reviewing Editor; University of the Witwatersrand, South Africa
  2. Helena Boshoff
    Reviewer; National Institute of Allergy and Infectious Diseases, United States

Our editorial process produces two outputs: (i) public reviews designed to be posted alongside the preprint for the benefit of readers; (ii) feedback on the manuscript for the authors, including requests for revisions, shown below. We also include an acceptance summary that explains what the editors found interesting or important about the work.

Decision letter after peer review:

Thank you for submitting your article "Pleiotropic roles for mycobacterial DinB2 in frameshift and substitution mutagenesis" for consideration by eLife. Your article has been reviewed by 3 peer reviewers, and the evaluation has been overseen by a Reviewing Editor and Bavesh Kana as the Senior Editor. The following individual involved in the review of your submission has agreed to reveal their identity: Helena Boshoff (Reviewer #2).

The reviewers have discussed their reviews with one another, and the Reviewing Editor has drafted this to help you prepare a revised submission.

Essential revisions:

1. Whilst the results are interesting, overexpression of the different DinBs is an artificial way of looking at their biological roles. Just because they increase mutation frequency when overproduced doesn't mean they act in DNA damage tolerance or mutagenesis when expressed at normal cellular levels from their native genomic loci. While this study is well done, it remains unclear whether DinB2 and/or DinB3 contribute to DNA damage tolerance and mutagenesis in Mycobacteria. Now that the authors have a unique mutation spectrum for DinB2 (when overexpressed), is it possible to search for that spectrum in response to DNA damage, and its disappearance in a strain lacking DinB2 function? This type of result would provide a strong argument that DinB2 is contributing to mutagenesis.

2. The dependence on the catalytic activity of overproduced levels of DinB2 to impede growth is very reminiscent of work from Walker (Foti et al., Science (2012) 336:315-319) concluding that overproduced levels of E. coli DinB (Pol IV) kill by incorporating oxidized precursors into nascent DNA, leading to lethal numbers of dsDNA breaks. Have the authors tested whether killing induced by the overproduction of DinB2 can be suppressed by anaerobic growth or similar treatments to limit the production of oxidized DNA precursors?

3. There are several concerns with the Western blots. First, can the authors comment on why panels 1A and 1E look so different? Second, quantitation of the blots would be helpful to gauge the expression levels of the DinBs? For example, was DinB3 expressed at the same level as DinB2, as a function of ATc concentration, and is growth impairment correlated with levels of DinB2 protein and not just ATc? Third, how many times were blots performed? Assuming each represents one of multiple reps their quantitation could indicate the average value +/- error (range or SD). Fourth, by eye, it looks like the DinB2L14F mutant (Figure 5) may be expressed at a slightly lower level than WT DinB2, but it has a more pronounced killing phenotype. Quantitation would support or refute the possibility that the ability to incorporate rNTPs impacts the growth defect caused by DinB2 overproduction. Finally, in Figure 6 was it confirmed that variable metal (Mg, Mn) concentrations failed to influence DinB2 levels by Western blot analysis?

4. The finding of kanR CFUs without any detectable mutations in the kan marker is worrisome and should be better discussed in the text. The same for sacB data in the supplementary material. The explanation given in lines 216-218 does not make sense. Markers 7G and 8G clearly are barely measuring any mutagenesis. The experiments wherein most of the supposed KanR revertants have no Kan mutation should either be removed from the manuscript or better discussed, because it is uncertain what they are measuring, therefore no conclusion can be drawn from them. For the Kan markers, one possible explanation is that translational frameshifts are occurring and allow residual growth of some of the cells. Gene amplification as seen in the lac system of Cairns and Foster in E. coli could also promote growth without actual mutations. Is the KanR phenotype of these colonies heritable and stable?

5. Also, spontaneous mutagenesis should have been more precisely measured by using fluctuation analysis of larger sample sizes. In many instances, the results shown are the means of a few cultures with very large differences in mutant frequencies (several hundred-fold – e.g. Figures 4C,D and E, 5C and F, S3). Authors could discuss/explain their choice of statistical analysis and sample sizes.

Other issues that should be addressed:

1. Title: Why are the authors are claiming "pleiotrpic" roles for these polymerases? What are the multiple phenotypes being observed, besides an effect on mutagenesis? The abstract is also not clear about this.

2. Line 44: Error in the citation of Boshoff 2003.

3. Figures 1B and 1F: How OD600 values are measured? Are these very high densities really achievable in mycobacterial cultures (OD600 = 100)?

4. A simple change that would add a lot to the paper. In all gel images showing polymerase activity, an indicator of the molecular size of the relevant products discussed in the text would be of tremendous help to the readers. It is often confusing to follow what is being discussed about the length of slippage tracts without this aid.

5. Lines 169-180: Why do the authors interpret products with an extra 8nt as the result of 8 rounds of slippage? Although this is possible, one slippage event could loop out more than one nucleotide, giving rise to the addition of more than one nucleotide. This idea is repeated later in the text when discussing figure 4.

6. Line 180: "single cycle of forward slippage".

7. Line 249: Klenow polymerase domain.

8. Line 252: "DinB2 slippage in G or C…"

9. Line 259 and 260. The difference observed in the Kan marker 5G is not significant, as indicated in the figure.

10. Lines 333 to 334: "Magnesium did not impact the relatively high level of slippage on the G6 template that was seen in the presence of ddATP." This is not clear, given that the experiment shows a clear effect of Mn ions. Would the point be that increasing concentrations of Mn do not enhance slippage significantly after the transition seen at 0.5 mM?

11. Figures 6E and 6F: indicate which reporters are being used. 4T, 5T?

12. The authors state in Methods that at least 3 cultures were sequenced for each mutagenesis experiment, but how many of the sequences came from each culture?

13. Figure S2 shows results for only DinB2, not DinB2 and DinB3, as stated in the text (lines 104-107).

14. Did the authors examine DinB3 replication on some of the same templates as DinB2, and if so did it also slip? If not, this would further support the conclusion that slippage is the cause of the frameshifts.

15. Line 250: was this slippage or misincorporation?

16. Both DinB2 and DinB3 over-expression result in the induction of RecA. The extent of upregulation of the SOS response is not readily discerned from the RecA immunoblot. Do the authors have any associated transcriptional data to investigate any differences?

Figure S2: the DinB3 over-expression in the dnaE2 and recA deletion mutants is missing but referred to in the text. This is an important control and the panel would be good to include.

17. Manganese supplementation of the culture medium does not increase the frequency of slippage in the induced tet-dinB2 strain and probably does not significantly affect that in the control strain either. A good additional control would have been the mutant lacking dnaE2 and dinB1-3 used in the Dupuy et al. (2022) study.

18. This study demonstrates the role of DinB2-mediated mutagenesis in vivo. It can incorporate ribonucleotides as well as dNTPs. The ratio of these could affect its in vivo activity. One wonders whether the over-expression of DinB2 and the DinB2-L14F (steric mutant) result in altered mutation frequencies when plated from starved cells.

19. Figure 1A vs Figure 5A: why is the DinB immunoblotting so weak in 1A?

20. Line 423: it would be informative to compare homo-oligonucleotide runs in bacteria such as E. coli and B. subtilis to get a better sense of how common such tracts are between different species.

Reviewer #1 (Recommendations for the authors):

This is an interesting study that follows up on a recent important publication from the same group (Dupuy et al., Nature Communications (2022) 13:4493). The M. smegmatis and M. tuberculosis DinB proteins (DinB1-DinB3 for Msm; DinB1-DinB2 for Mtb) were previously thought to play no role in DNA damage tolerance and mutagenesis. However, the recent work from Dupuy et al. shows that DinB1 from both organisms plays important roles in DNA damage-induced mutagenesis. By contrast, the authors failed to observe a phenotype for DinB2 and DinB3 mutants. In the current work, the authors have studied the impact of DinB2 and DinB3 overproduction on mutation frequency/signature, concluding that DinB2 plays important roles in both substitution and frameshift mutagenesis and that Mn contributes to the mutator phenotype. Finally, they convincingly demonstrate that frameshifts are the result of DinB2 slipping in homopolymeric nucleotide runs during dNTP and not rNTP incorporation. With the exceptions noted below, this work is well-designed, well-performed, and well-presented. The frameshift results are a strength and represent the most complete and systematic analysis this reviewer has seen. The main concern with the work is that it has studied DinB2 and DinB3 under very artificial conditions that may not represent biologically important (or even relevant) roles for these proteins in M. smegmatis. As such, it remains to be demonstrated whether DinB2 and/or DinB3 contribute to DNA damage tolerance and mutagenesis in vivo when expressed at normal cellular levels from their native genomic loci. This and other concerns the authors should address are discussed below.

1) I appreciate why the authors did what they did, and while I agree their results are interesting, overexpression of the different DinBs is an artificial way of looking at their biological roles. Just because they increase mutation frequency when overproduced doesn't mean they act in DNA damage tolerance or mutagenesis when expressed at normal cellular levels from their native genomic loci. While this study is well done, it remains unclear whether DinB2 and/or DinB3 contribute to DNA damage tolerance and mutagenesis in Mycobacteria. Now that the authors have a unique mutation spectrum for DinB2 (when overexpressed), is it possible to search for that spectrum in response to DNA damage, and its disappearance in a strain lacking DinB2 function? This type of result would provide a strong argument that DinB2 is contributing to mutagenesis.

2) The dependence on catalytic activity of overproduced levels of DinB2 to impede growth is very reminiscent of work from Walker (Foti et al., Science (2012) 336:315-319) concluding that overproduced levels of E. coli DinB (Pol IV) kill by incorporating oxidized precursors into nascent DNA, leading to lethal numbers of dsDNA breaks. Have the authors tested whether killing induced by the overproduction of DinB2 can be suppressed by anaerobic growth or similar treatments to limit the production of oxidized DNA precursors?

3) There are several concerns with the Western blots. First, can the authors comment on why panels 1A and 1E look so different? Second, quantitation of the blots would be helpful so the reader knew at what levels the DinBs were expressed. For example, was DinB3 expressed at the same levels as DinB2 as a function of ATc concentration, and is growth impairment correlated with levels of DinB2 protein and not just ATc? Third, how many times were blots performed? Assuming each represents one of multiple reps their quantitation could indicate the average value +/- error (range or SD). Fourth, by eye, it looks like the DinB2L14F mutant (Figure 5) may be expressed at a slightly lower level than WT DinB2, but it has a more pronounced killing phenotype. Quantitation would support or refute the possibility that the ability to incorporate rNTPs impacts the growth defect caused by DinB2 overproduction. Finally, in Figure 6 was it confirmed that variable metal (Mg, Mn) concentrations failed to influence DinB2 levels by Western blot analysis?

4) Several of the gels showing DinB2 replication products are lacking size standards, making it very difficult to follow along with the discussion of the results.

Reviewer #3 (Recommendations for the authors):

Title: I do not follow why the authors are claiming "pleiotrpic" roles for these polymerases. What are the multiple phenotypes being observed, besides an effect on mutagenesis? The abstract is also not clear about this.

Line 44: Error in the citation of Boshoff 2003.

Figures 1B and 1F: How OD600 values are measured? Are these very high densities really achievable in mycobacterial cultures (OD600 = 100)?

A simple change that would add a lot to the paper. In all gel images showing polymerase activity, an indicator of the molecular size of the relevant products discussed in the text would be of tremendous help to the readers. It is often confusing to follow what is being discussed about the length of slippage tracts without this aid.

Lines 169-180: I did not understand why authors interpret products with an extra 8nt as the result of 8 rounds of slippage. Although this is possible, one slippage event could loop out more than one nucleotide, giving rise to the addition of more than one nucleotide. This idea is repeated later in the text when discussing figure 4.

Line 180: "single cycle of forward slippage".

Line 249: Klenow polymerase domain.

Line 252: "DinB2 slippage in G or C…"

Line 259 and 260. The difference observed in the Kan marker 5G is not significant, as indicated in the figure.

Lines 333 to 334: "Magnesium did not impact the relatively high level of slippage on the G6 template that was seen in the presence of ddATP.". I did not understand the point raised here, given that the experiment shows a clear effect of Mn ions. Would the point be that increasing concentrations of Mn do not enhance slippage significantly after the transition seen at 0.5 mM?

Figure 6 E and 6F: indicate which reporters are being used. 4T, 5T?

[Editors' note: further revisions were suggested prior to acceptance, as described below.]

Thank you for resubmitting your work entitled "Roles for mycobacterial DinB2 in frameshift and substitution mutagenesis" for further consideration by eLife. Your revised article has been evaluated by Bavesh Kana (Senior Editor) and a Reviewing Editor.

The manuscript has been improved but there are some remaining issues that need to be addressed, as outlined below:

Reviewers agreed that the revised manuscript has been substantively improved, there is however one primary concern remaining. This relates to ascribing a function for DinB2 in mutagenesis, particularly the conditions under which such activity manifests. As a first request, please consider revising your discussion, and appending text in the abstract to specifically address this limitation. Refer to the detailed comments from the reviewers to get a better sense of this concern, particularly point 1 from Reviewer 1.

Secondly, the following points need addressing, also through revision of the manuscript to acknowledge any limitations and clarify data presentation:

1. The concern about the variation between western blots in panels 1A vs 1E referred to the variable signal for the RpoD loading control. The weak RpoD signal (in contrast to other RpoD panels) may be outside the linear range of detection, drawing into question the accuracy of its use as a loading control and normalizing standard for the quantitation of RecA. Likewise, in panel A of Figure 1-supplement figure 1, the signal for the DinB2 and DinB3 proteins was very low (in contrast to other DinB2/DinB3 panels) and may not be in the linear range of detection. Can you the authors explain this, specifically with respect to the accuracy of the quantitation of these low signals.

2. Figure 4, supplemental panels 1A-D, and figure 4B would benefit from more labels of molecular sizes for DNA products. For example, in figure 4 the authors state that products were in excess "~+60" in size, but the size markings only go up to 50 bp – it would greatly help the reader to see where the +60 product migrates to.

3. Pg. 6, lines 129-133: DinB1 is regulated by SigH, but can the authors exclude a role for chromosomally-expressed DinB1 in contributing to mutagenesis, similar to the arguments for RecA (SOS) and DnaE2?

4. Pg. 7, lines 158-162: The authors may want to relax/qualify their statement regarding DinB3 as they did not use a DinB3 active site mutation to confirm that mutations observed were in fact the result of DinB3 catalytic activity.

Reviewer #1 (Recommendations for the authors):

Recommendations for the authors: While I enjoyed reading this paper, and this revised version is improved, I do have a few comments, some that were raised in the initial review and were not completely addressed and some that are new and are in response to some of the conclusions stated in the revised version.

1) "Essential revision" point #1: a remaining concern is whether levels of the DinB2/B3 proteins examined in this work are higher than when induced from the native genomic loci using, for example, menaquinone to induce dinB2 (I don't believe they discussed the ability of menaquinone to induce dinB2 in their initial draft). I think this point is important because overexpression of a DNA polymerase can confer phenotypes that have nothing to do with its true biological function. For example, the Sweasy lab developed genetic selections and screens for rat DNA polymerase β (Pol β) overexpressed in E. coli based on the ability of Pol β to access DNA in a recA polA mutant (e.g., PNAS, 94: 321-1326 and JBC, 274(6): 3851-3858). Thus, while it may be the case, as the authors suggest, that failure to observe a phenotype for the DinB2 mutant may be the result of the conditions studied failing to induce DinB2 (or DinB3) expression, it may also be the case that DinB2/DinB3 fail to confer a mutator phenotype in Msm unless expressed at higher than chromosomally expressed level. Conditions for induction of DinB2 expression are known (e.g., menaquinone, ∆MSMEG_2294, etc.) – why not try these conditions to induce physiologically relevant levels of chromosomally-expressed DinB2 and look for mutator phenotypes? I am still concerned that the in vivo mutagenic properties described here using overexpressed levels of DinB2/DinB3 may not be biologically relevant.

2) "Essential revision" point #3: the concern about the variation between WB in panels 1A vs 1E referred to the variable signal for the RpoD loading control. The weak RpoD signal (in contrast to other RpoD panels) may be outside the linear range of detection, drawing into question the accuracy of its use as a loading control and normalizing standard for the quantitation of RecA. Likewise, in panel A of Figure 1-supplement figure 1, the signal for the DinB2 and DinB3 proteins was very low (in contrast to other DinB2/DinB3 panels) and may not be in the linear range of detection. I am curious about the authors' views on why these panels have low signals while other panels do not, and the accuracy of the quantitation of these low signals.

3) "Other issues that should be addressed" point #4: figure 4, supplemental panels 1A-D, and figure 4B would benefit from more labels of molecular sizes for DNA products. For example, in figure 4 the authors state that products were in excess "~+60" in size, but the size markings only go up to 50 bp – it would greatly help the reader to see where the +60 product migrates to.

4) Pg. 6, lines 129-133: I realize that DinB1 is regulated by SigH, but can the authors exclude a role for chromosomally-expressed DinB1 in contributing to mutagenesis, similar to the arguments for RecA (SOS) and DnaE2?

5) Pg. 7, lines 158-162: The authors may want to relax/qualify their statement regarding DinB3 as they did not use a DinB3 active site mutation to confirm that mutations observed were in fact the result of DinB3 catalytic activity.

Reviewer #2 (Recommendations for the authors):

The authors have sufficiently addressed the reviewers' concerns. The authors explore the catalytic activity and mutagenic profile of DinB2. The intracellular function of DinB2 under physiological conditions remains elusive since only over-expression studies hint at its role in mutagenesis. Nevertheless, DinB2 may play a discernable role under certain conditions which upregulate its expression.

https://doi.org/10.7554/eLife.83094.sa1

Author response

Essential revisions:

1. Whilst the results are interesting, overexpression of the different DinBs is an artificial way of looking at their biological roles. Just because they increase mutation frequency when overproduced doesn't mean they act in DNA damage tolerance or mutagenesis when expressed at normal cellular levels from their native genomic loci. While this study is well done, it remains unclear whether DinB2 and/or DinB3 contribute to DNA damage tolerance and mutagenesis in Mycobacteria. Now that the authors have a unique mutation spectrum for DinB2 (when overexpressed), is it possible to search for that spectrum in response to DNA damage, and its disappearance in a strain lacking DinB2 function? This type of result would provide a strong argument that DinB2 is contributing to mutagenesis.

We agree that our work does not demonstrate that DinB2 has mutagenic activities in physiological conditions. This paper is a characterization of mutagenic proprieties of DinB2 in vivo and in vitro as well as homo-oligonucleotide runs. In our previous paper (PMID: 35918328), we found that the dinB2 deletion does not cause a decrease of mutation frequency (substitutions or indels) and a focus on DinB2-specific mutations also did not reveal an effect. However, we note that the absence of an effect of a loss of function approach is only interpretable if the gene product is expressed in the conditions tested. There is evidence that dinB2 expression is induced only under specific conditions. For instance, Barman et al., 2022 revealed that Menaquinone induces dinB2 expression more than 10-fold. Although we cannot compare the absolute levels of DinB2 in our experiments to that study (due to their use of mRNA and our use of an epitope tagged protein) we have quantitated replicate western blots (as requested), which show a 10-fold increase of protein DinB2 level between – and + ATc conditions (Figure 1A), similar to the induction ratio in the literature. Thus, although we agree that our study determines the mutagenic potential of DinB2 without defining its physiologic role, the lack of a phenotype of the DinB2 deletion strain can be explained by lack of expression. We have added this reasoning (lines 66-72) to the manuscript introduction to justify our use of the inducible expression system.

2. The dependence on the catalytic activity of overproduced levels of DinB2 to impede growth is very reminiscent of work from Walker (Foti et al., Science (2012) 336:315-319) concluding that overproduced levels of E. coli DinB (Pol IV) kill by incorporating oxidized precursors into nascent DNA, leading to lethal numbers of dsDNA breaks. Have the authors tested whether killing induced by the overproduction of DinB2 can be suppressed by anaerobic growth or similar treatments to limit the production of oxidized DNA precursors?

We agree that the hypothesis by which the lethality induced by dinB2 overexpression is due to genomic incorporation/excision of 8-oxoguanines (8-oxoG), as demonstrated in E. coli by Foti et al., is very attractive. To test this possibility, we measured the effect of dinB2 OE on bacterial growth and viability in mutT1234 (8-oxoG degradation systems) and mutYM12 (8-oxoG excision systems) mutants. We did not observe significant differences between genetic backgrounds suggesting that DinB2-dependent lethality is not mainly due to 8-oxoG genomic incorporation/excision.

This paragraph was added (lines 95-104) in addition to Figure1—figure supplement 2: “A study conducted in E. coli revealed that dinB overexpression toxicity is due to genomic incorporation and excision of 8-oxoguanines, leading to the formation of DNA double-strand breaks (Foti et al., 2012). This is reminiscent with the ability of DinB2 to utilize 8-oxoguanine for DNA synthesis in vitro (Ordonez & Shuman, 2014). To test if DinB2 overexpression causes cell death in vivo because of genomic incorporation of 8-oxoguanines, we measured the impact of mutTs and mutY/mutMs deletions on DinB2-dependant lethality, systems involved in free 8-oxoguanines degradation and genomic 8-oxoguanines excision, respectively (Dupuy et al., 2020). We did not observe a significant impact of the absence of these 8-oxoguanine processing systems on the effect of DinB2 (Figures S2A- D), indicating that DinB2-dependent lethality is not mainly due to genomic 8-oxoguanine incorporation.”

3. There are several concerns with the Western blots. First, can the authors comment on why panels 1A and 1E look so different? Second, quantitation of the blots would be helpful to gauge the expression levels of the DinBs? For example, was DinB3 expressed at the same level as DinB2, as a function of ATc concentration, and is growth impairment correlated with levels of DinB2 protein and not just ATc? Third, how many times were blots performed? Assuming each represents one of multiple reps their quantitation could indicate the average value +/- error (range or SD). Fourth, by eye, it looks like the DinB2L14F mutant (Figure 5) may be expressed at a slightly lower level than WT DinB2, but it has a more pronounced killing phenotype. Quantitation would support or refute the possibility that the ability to incorporate rNTPs impacts the growth defect caused by DinB2 overproduction. Finally, in Figure 6 was it confirmed that variable metal (Mg, Mn) concentrations failed to influence DinB2 levels by Western blot analysis?

We thank the reviewer for this suggestion. In response, we conducted new western blots (WB) of biological triplicates that are now included:

1. WB of DinB2 and DinB3 with different concentrations of ATC (Figure 1A).

2. WB of DinB2, DinB2L107A, and DinB2L14F (Figure 1E and 5A).

3. WB of DinB2 in presence of variable concentrations of MnCl2 (Figure 6C).

We also performed quantifications of RecA WB presented in the 1st version of the paper. The number of biological replicates and protein quantifications are now indicated under the picture of a representative blot.

4. The finding of kanR CFUs without any detectable mutations in the kan marker is worrisome and should be better discussed in the text. The same for sacB data in the supplementary material. The explanation given in lines 216-218 does not make sense. Markers 7G and 8G clearly are barely measuring any mutagenesis. The experiments wherein most of the supposed KanR revertants have no Kan mutation should either be removed from the manuscript or better discussed, because it is uncertain what they are measuring, therefore no conclusion can be drawn from them. For the Kan markers, one possible explanation is that translational frameshifts are occurring and allow residual growth of some of the cells. Gene amplification as seen in the lac system of Cairns and Foster in E. coli could also promote growth without actual mutations. Is the KanR phenotype of these colonies heritable and stable?

We agree with the reviewer that the 7G and 8G kanR reporters are not contributing to the study and we have removed them.

For the results obtained with SacB reporters, in the control strain (empty vector), a majority of sucR clones do not have a sacB mutation, but their proportion is strongly induced after dinB2 OE (because the frequency of FS in sacB increases). In addition to supporting the +1 and -1 FS activity of DinB2 in 6C and 9C runs, the primary purpose of the sacB data, because of its ability to report on longer insertion events due to its counter selectable function, is to exclude longer insertions that would be invisible in the kan system. For these two reasons, we have retained the sacB reporters in Figure 4—figure supplement 2.

5. Also, spontaneous mutagenesis should have been more precisely measured by using fluctuation analysis of larger sample sizes. In many instances, the results shown are the means of a few cultures with very large differences in mutant frequencies (several hundred-fold – e.g. Figures 4C,D and E, 5C and F, S3). Authors could discuss/explain their choice of statistical analysis and sample sizes.

We agree that fluctuation analysis is a very good way to compare spontaneous mutation frequencies between strains, cultivated through many generations because it corrects for jackpot effects of mutations arising early in the growth experiment. In our study, we measured mutation frequency after a short induction of DinBs using a 16h of ATC treatment (around 6 generations only), limiting the risk of jackpot effect. Because we observed a very strong induction of the mutation frequency (between 10- and 100-fold induction for most of the runs), and given the large number of strains/conditions tested in our study, we believe that the measure of mutation frequencies rather than fluctuation test (that needs several tens of replicates per condition) is appropriate.

Other issues that should be addressed:

1. Title: Why are the authors are claiming "pleiotrpic" roles for these polymerases? What are the multiple phenotypes being observed, besides an effect on mutagenesis? The abstract is also not clear about this.

Pleiotropic has been deleted from the title.

2. Line 44: Error in the citation of Boshoff 2003.

Corrected.

3. Figures 1B and 1F: How OD600 values are measured? Are these very high densities really achievable in mycobacterial cultures (OD600 = 100)?

These growth curves, as stated in the methods, are continuous growth experiments to deduce an accurate doubling time. Cultures are diluted in fresh medium when they reach an OD600=1 and subsequent measured OD600 are corrected by the dilution factor for plotting. Thus, the Y axis does not represent a true OD, it is a calculated number. The procedure is described in the Methods section: “OD600 was measured for 48h. Cultures were diluted in fresh medium supplemented with ATc to a calculated OD600=0.001 when OD600 reached a value around 1 and measured OD600 was corrected by the dilution factor.”.

4. A simple change that would add a lot to the paper. In all gel images showing polymerase activity, an indicator of the molecular size of the relevant products discussed in the text would be of tremendous help to the readers. It is often confusing to follow what is being discussed about the length of slippage tracts without this aid.

Sizes (in nucleotides) of the relevant products are now indicated in the Figures depicting polymerase assays.

5. Lines 169-180: Why do the authors interpret products with an extra 8nt as the result of 8 rounds of slippage? Although this is possible, one slippage event could loop out more than one nucleotide, giving rise to the addition of more than one nucleotide. This idea is repeated later in the text when discussing figure 4.

We have added the following text:

“The heterogeneous size distribution of the slippage ladder is consistent with either of two scenarios: (i) multiple slippage cycles in which the primer 3’-OH end realigns backward on the template by a single nucleotide; or (ii) one or several cycles of backward realignment of the primer 3’-OH on the template by more than one nucleotide (the upper limit being the length of the template homooligomeric tract) followed by fill-in to the end of the homooligomeric tract.”

6. Line 180: "single cycle of forward slippage".

Corrected.

7. Line 249: Klenow polymerase domain.

Corrected.

8. Line 252: "DinB2 slippage in G or C…"

Corrected.

9. Line 259 and 260. The difference observed in the Kan marker 5G is not significant, as indicated in the figure.

The sentence was changed for: “However, dinB2 overexpression enhanced -1 FS frequency in the 4G run by 100-fold and non-statistically significant increase of +1 FS frequency was observed in the 5G run.”

10. Lines 333 to 334: "Magnesium did not impact the relatively high level of slippage on the G6 template that was seen in the presence of ddATP." This is not clear, given that the experiment shows a clear effect of Mn ions. Would the point be that increasing concentrations of Mn do not enhance slippage significantly after the transition seen at 0.5 mM?

The metal concentration dependence of the shift in slippage behavior is relevant insofar as the standard assays for slippage contained 5 mM manganese (which we now state explicitly in Results text) and the experiment presented makes clear that the shift in the presence of 5 mM magnesium occurs at lower concentrations of manganese (0.5 mM).

11. Figures 6E and 6F: indicate which reporters are being used. 4T, 5T?

Done.

12. The authors state in Methods that at least 3 cultures were sequenced for each mutagenesis experiment, but how many of the sequences came from each culture?

It differs depending on the experiment. the sentence “(between 2 and 8 clones per independent culture were sequenced)” was added to the methods section.

13. Figure S2 shows results for only DinB2, not DinB2 and DinB3, as stated in the text (lines 104-107).

DinB3 data was added to Figure 2—figure supplement 1.

14. Did the authors examine DinB3 replication on some of the same templates as DinB2, and if so did it also slip? If not, this would further support the conclusion that slippage is the cause of the frameshifts.

Although we agree this is an interesting question, the in vitro frameshifting activity of DinB3 was not examined here.

15. Line 250: was this slippage or misincorporation?

We cannot distinguish, but the noted activity of PolI is much weaker than DinB2.

16. Both DinB2 and DinB3 over-expression result in the induction of RecA. The extent of upregulation of the SOS response is not readily discerned from the RecA immunoblot. Do the authors have any associated transcriptional data to investigate any differences?

Figure S2: the DinB3 over-expression in the dnaE2 and recA deletion mutants is missing but referred to in the text. This is an important control and the panel would be good to include.

Data of substitution mutagenesis after dinB3 OE in ΔrecA and ΔdnaE2 mutants were added in Figure 2—figure supplement 1.

17. Manganese supplementation of the culture medium does not increase the frequency of slippage in the induced tet-dinB2 strain and probably does not significantly affect that in the control strain either. A good additional control would have been the mutant lacking dnaE2 and dinB1-3 used in the Dupuy et al. (2022) study.

We are not clear on what the reviewer is asking. The lack of effect of Mn in the control strain would seem to exclude the involvement of any of the polymerases and therefore the null strains would not be informative.

18. This study demonstrates the role of DinB2-mediated mutagenesis in vivo. It can incorporate ribonucleotides as well as dNTPs. The ratio of these could affect its in vivo activity. One wonders whether the over-expression of DinB2 and the DinB2-L14F (steric mutant) result in altered mutation frequencies when plated from starved cells.

We agree with this point by the reviewer and exploring the role of DinB2 in starved cells will be part of a future study.

19. Figure 1A vs Figure 5A: why is the DinB immunoblotting so weak in 1A?

New western-blots were performed in triplicate and band intensity was quantified as noted above and in the figures.

20. Line 423: it would be informative to compare homo-oligonucleotide runs in bacteria such as E. coli and B. subtilis to get a better sense of how common such tracts are between different species.

We agree that this is an interesting bioinformatic question and we agree that our study and others may stimulate such a comparison.

[Editors' note: further revisions were suggested prior to acceptance, as described below.]

The manuscript has been improved but there are some remaining issues that need to be addressed, as outlined below:

Reviewers agreed that the revised manuscript has been substantively improved, there is however one primary concern remaining. This relates to ascribing a function for DinB2 in mutagenesis, particularly the conditions under which such activity manifests. As a first request, please consider revising your discussion, and appending text in the abstract to specifically address this limitation. Refer to the detailed comments from the reviewers to get a better sense of this concern, particularly point 1 from Reviewer 1.

Secondly, the following points need addressing, also through revision of the manuscript to acknowledge any limitations and clarify data presentation:

1. The concern about the variation between western blots in panels 1A vs 1E referred to the variable signal for the RpoD loading control. The weak RpoD signal (in contrast to other RpoD panels) may be outside the linear range of detection, drawing into question the accuracy of its use as a loading control and normalizing standard for the quantitation of RecA. Likewise, in panel A of Figure 1-supplement figure 1, the signal for the DinB2 and DinB3 proteins was very low (in contrast to other DinB2/DinB3 panels) and may not be in the linear range of detection. Can you the authors explain this, specifically with respect to the accuracy of the quantitation of these low signals.

We presume these comments relate to RpoB, the loading control used, not RpoD. Regarding the linear range of detection, as stated in the methods, these blots were imaged on an iBright FL1000, which has a large dynamic range and alerts the user to bands that are outside of the dynamic range of the instrument. We can confirm that this quantitation was done in the linear range and have added this detail to the methods. Regarding the statement that the band intensity is different between gels, we emphasize that variable band intensity between blots is intrinsic to the technique of immunoblotting. Variation in transfer efficiency, antibody variability, age of chemiluminescent reagent, and many other variable affect band intensities between blots. This variability is exactly the reason that we (and others) employ loading controls as the most rigorous way to normalize these variables and provide an internal control for quantitation on each blot, which we have done for 3 replicate blots to derive the numbers that are presented.

2. Figure 4, supplemental panels 1A-D, and figure 4B would benefit from more labels of molecular sizes for DNA products. For example, in figure 4 the authors state that products were in excess "~+60" in size, but the size markings only go up to 50 bp – it would greatly help the reader to see where the +60 product migrates to.

Unfortunately, we are unable to provide additional markers on these gels. We have altered the text to say >50 since the products are above the 50bp marker.

3. Pg. 6, lines 129-133: DinB1 is regulated by SigH, but can the authors exclude a role for chromosomally-expressed DinB1 in contributing to mutagenesis, similar to the arguments for RecA (SOS) and DnaE2?

We thank for reviewer for this comment. The mutation spectrum catalyzed by DinB2 is most similar to DnaE2 in that it appears in multiple codons within the RRDR. As noted in the text, the RifR mutations catalyzed by DinB1 (and DinB3, as shown here) are focused on a single substitution with a single codon (His442 CAC>CGC). Although this mutation is part of the DinB2 mutation spectrum (see figure 4D) it is a small fraction and therefore it is unlikely that DinB1 is responsible for the DinB2 mutation picture. We cannot exclude that DinB1 contributes to the small number of His442 CAC>CGC mutations observed.

4. Pg. 7, lines 158-162: The authors may want to relax/qualify their statement regarding DinB3 as they did not use a DinB3 active site mutation to confirm that mutations observed were in fact the result of DinB3 catalytic activity.

We have added the following sentence to the discussion:

“This DinB2 effect is dependent on its polymerase activity, whereas we did not express a DinB3 active site mutant.”

https://doi.org/10.7554/eLife.83094.sa2

Article and author information

Author details

  1. Pierre Dupuy

    Immunology Program, Sloan Kettering Institute, New York, United States
    Contribution
    Conceptualization, Formal analysis, Investigation, Methodology, Writing – original draft, Writing – review and editing
    Contributed equally with
    Shreya Ghosh
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-7451-304X

  2. Shreya Ghosh

    Molecular Biology Program, Sloan Kettering Institute, New York, United States
    Contribution
    Conceptualization, Formal analysis, Investigation, Methodology, Writing – review and editing
    Contributed equally with
    Pierre Dupuy
    Competing interests
    No competing interests declared

  3. Allison Fay

    Immunology Program, Sloan Kettering Institute, New York, United States
    Contribution
    Investigation
    Competing interests
    No competing interests declared

  4. Oyindamola Adefisayo

    1. Immunology Program, Sloan Kettering Institute, New York, United States
    2. Immunology and Microbial Pathogenesis Graduate Program, Weill Cornell Graduate School, New York, United States
    Contribution
    Conceptualization, Investigation
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0001-9376-148X

  5. Richa Gupta

    Immunology Program, Sloan Kettering Institute, New York, United States
    Contribution
    Conceptualization, Investigation, Methodology
    Competing interests
    No competing interests declared

  6. Stewart Shuman

    Molecular Biology Program, Sloan Kettering Institute, New York, United States
    Contribution
    Conceptualization, Resources, Supervision, Funding acquisition, Investigation, Writing – original draft, Writing – review and editing
    For correspondence
    shumans@mskcc.org
    Competing interests
    No competing interests declared

  7. Michael S Glickman

    1. Immunology Program, Sloan Kettering Institute, New York, United States
    2. Immunology and Microbial Pathogenesis Graduate Program, Weill Cornell Graduate School, New York, United States
    Contribution
    Conceptualization, Resources, Supervision, Funding acquisition, Investigation, Writing – original draft, Writing – review and editing
    For correspondence
    glickmam@mskcc.org
    Competing interests
    MG has received consulting fees from Vedanta Biosciences, PRL NYC, and Fimbrion Therapeutics and has equity in Vedanta biosciences
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0001-7918-5164

Funding

National Institute of Allergy and Infectious Diseases (AI064693)

  • Stewart Shuman
  • Michael S Glickman

National Cancer Institute (P30CA008748)

  • Pierre Dupuy
  • Michael S Glickman
  • Stewart Shuman
  • Shreya Ghosh
  • Allison Fay
  • Oyindamola Adefisayo
  • Richa Gupta

Institut National de la Recherche Agronomique (Département Caractérisation et Élaboration des Produits Issus de l'Agriculture: Jeune scientifique)

  • Pierre Dupuy

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Acknowledgements

This work is supported by NIH (NIH grant #AI064693) and this research was funded in part through the NIH/NCI Cancer Center Support Grant P30CA008748. P Dupuy was supported in part by a 'Jeune Scientifique' salary award from the French National Institute of Agronomic Science (INRA). We thank all Glickman and Shuman lab members for helpful discussions.

Senior and Reviewing Editor

  1. Bavesh D Kana, University of the Witwatersrand, South Africa

Reviewer

  1. Helena Boshoff, National Institute of Allergy and Infectious Diseases, United States

Publication history

  1. Preprint posted: August 14, 2022 (view preprint)
  2. Received: August 31, 2022
  3. Accepted: April 18, 2023
  4. Version of Record published: May 4, 2023 (version 1)

Copyright

© 2023, Dupuy, Ghosh et al.

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.

Metrics

  • 143
    Page views
  • 29
    Downloads
  • 0
    Citations

Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. Pierre Dupuy
  2. Shreya Ghosh
  3. Allison Fay
  4. Oyindamola Adefisayo
  5. Richa Gupta
  6. Stewart Shuman
  7. Michael S Glickman
(2023)
Roles for mycobacterial DinB2 in frameshift and substitution mutagenesis
eLife 12:e83094.
https://doi.org/10.7554/eLife.83094

Categories and tags

Further reading

    1. Immunology and Inflammation
    2. Microbiology and Infectious Disease

    Cellular and molecular dynamics in the lungs of neonatal and juvenile mice in response to E. coli

    Sharon A McGrath-Morrow, Jarrett Venezia ... Alan L Scott
    Research Article

    Bacterial pneumonia in neonates can cause significant morbidity and mortality when compared to other childhood age groups. To understand the immune mechanisms that underlie these age-related differences, we employed a mouse model of E. coli pneumonia to determine the dynamic cellular and molecular differences in immune responsiveness between neonates (PND 3-5) and juveniles (PND 12-18), at 24, 48, and 72 hours. Cytokine gene expression from whole lung extracts was also quantified at these time points, using qRT-PCR. E. coli challenge resulted in rapid and significant increases in neutrophils, monocytes, and γδT cells, along with significant decreases in dendritic cells and alveolar macrophages in the lungs of both neonates and juveniles. E. coli challenged juvenile lung had significant increases in interstitial macrophages and recruited monocytes that were not observed in neonatal lungs. Expression of IFNg-responsive genes was positively correlated with the levels and dynamics of MHCII-expressing innate cells in neonatal and juvenile lungs. Several facets of immune responsiveness in the wild-type neonates were recapitulated in juvenile MHCII-/- juveniles. Employing a pre-clinical model of E. coli pneumonia, we identified significant differences in the early cellular and molecular dynamics in the lungs that likely contribute to the elevated susceptibility of neonates to bacterial pneumonia and could represent targets for intervention to improve respiratory outcomes and survivability of neonates.

    1. Microbiology and Infectious Disease

    Phage tRNAs evade tRNA-targeting host defenses through anticodon loop mutations

    Daan F van den Berg, Baltus A van der Steen ... Stan JJ Brouns
    Short Report

    tRNAs in bacteriophage genomes are widespread across bacterial host genera, but their exact function has remained unclear for more than 50 years. Several hypotheses have been proposed, and the most widely accepted one is codon compensation, which suggests that phages encode tRNAs that supplement codons that are less frequently used by the host. Here, we combine several observations and propose a new hypothesis that phage-encoded tRNAs counteract the tRNA-depleting strategies of the host using enzymes such as VapC, PrrC, Colicin D, and Colicin E5 to defend from viral infection. Based on mutational patterns of anticodon loops of tRNAs encoded by phages, we predict that these tRNAs are insensitive to host tRNAses. For phage-encoded tRNAs targeted in the anticodon itself, we observe that phages typically avoid encoding these tRNAs. Further supporting the hypothesis that phage tRNAs are selected to be insensitive to host anticodon nucleases. Altogether our results support the hypothesis that phage-encoded tRNAs have evolved to be insensitive to host anticodon nucleases.

    1. Microbiology and Infectious Disease

    The Plasmodium falciparum apicoplast cysteine desulfurase provides sulfur for both iron-sulfur cluster assembly and tRNA modification

    Russell P Swift, Rubayet Elahi ... Sean T Prigge
    Research Article Updated

    Iron-sulfur clusters (FeS) are ancient and ubiquitous protein cofactors that play fundamental roles in many aspects of cell biology. These cofactors cannot be scavenged or trafficked within a cell and thus must be synthesized in any subcellular compartment where they are required. We examined the FeS synthesis proteins found in the relict plastid organelle, called the apicoplast, of the human malaria parasite Plasmodium falciparum. Using a chemical bypass method, we deleted four of the FeS pathway proteins involved in sulfur acquisition and cluster assembly and demonstrated that they are all essential for parasite survival. However, the effect that these deletions had on the apicoplast organelle differed. Deletion of the cysteine desulfurase SufS led to disruption of the apicoplast organelle and loss of the organellar genome, whereas the other deletions did not affect organelle maintenance. Ultimately, we discovered that the requirement of SufS for organelle maintenance is not driven by its role in FeS biosynthesis, but rather, by its function in generating sulfur for use by MnmA, a tRNA modifying enzyme that we localized to the apicoplast. Complementation of MnmA and SufS activity with a bacterial MnmA and its cognate cysteine desulfurase strongly suggests that the parasite SufS provides sulfur for both FeS biosynthesis and tRNA modification in the apicoplast. The dual role of parasite SufS is likely to be found in other plastid-containing organisms and highlights the central role of this enzyme in plastid biology.