DNA polymerase IV in Escherichia coli

6 Affiliations: 7 1Molecular Horizons Institute and School of Chemistry and Molecular Bioscience, University of 8 Wollongong, Wollongong, Australia 9 2Illawarra Health and Medical Research Institute, Wollongong, Australia 10 3Department of Biochemistry, University of Wisconsin-Madison, USA 11 4Department of Biological Sciences, University of Southern California, Los Angeles, California, 12 United States of America 13 5Departments of Biological Sciences and Chemistry, University of Southern California, Los 14 Angeles, California, United States of America

In this work, we set out to test the following: 1. whether the UmuD cleavage status 111 affects the extent of pol IV focus formation at replisomes and/or the lifetimes of pol IV 112 molecules binding to DNA substrates, and 2. whether pol IV predominantly binds at RecA* 113 structures. Using single-molecule live-cell imaging, here we demonstrated that the DNA binding 114 activity of pol IV is promoted by UmuD in cells treated with the DNA damaging antibiotic 115 ciprofloxacin. In contrast, UmuD′ diminishes pol IV binding. We observed that a large 116 proportion of pol IV foci (up to 40%) colocalise with a RecA* marker. The recA(E38K) 117 mutation (also known as recA730), which constitutively produces RecA*-like activity [40][41][42], 118 promotes the binding activity of pol IV, even in the absence of DNA damage. We further showed 119 that this activity is due to pol IV interacting with RecA(E38K), which forms RecA*-like 120 structures on single-stranded as well as double-stranded DNA. These findings provide evidence 121 for regulatory roles for both UmuD and RecA in modulating the binding activity of pol IV in E. 122 coli cells. RecA* structures serve as major binding sites for pol IV, demonstrating a major role 123 for pol IV in homologous recombination.

125
Cleavage of UmuD affects the binding behaviour of pol IV 126 Previously, we showed that the limited colocalisation of pol IV foci with replisome 127 markers after treatment with DNA damaging agents drops from ~10% to < 5% (i.e. baseline To investigate the effect of UmuD status on pol IV activity, we compared colocalisation 133 between DinB-YPet and a DnaX-mKate2 replisome marker (coding for a fluorescent fusion of 134 the τ clamp loader protein, serving as a marker for the replisome, τ-mKate2) in four strains (all of 135 which include the dinB-YPet dnaX-mKate2 constructs): i) umuDC + (EAW643, [12]), ii) 136 umuDC (SSH007), iii) SSH007 expressing the non-cleavable UmuD(K97A) protein from a 137 low-copy plasmid (dinB-YPet dnaX-mKate2 umuDC + pUmuD[K97A], SSH007 + pJM1243), 138 and iv) SSH007 expressing the 'cleaved' UmuD′ protein from a low-copy plasmid (dinB-YPet 139 dnaX-mKate2 umuDC + pUmuD′, SSH007 + pRW66). The amount of UmuD(K97A) and 140 UmuD′ produced from each plasmid is 4-5-fold higher than UmuD expressed from its native 141 chromosomal locus [38,43]. 142 Time-lapse movies were recorded for each strain after treatment with ciprofloxacin 143 (30 ng mL -1 ). At t = 0 min, images of the DinB-YPet signal (300 ms exposure) and τ-mKate2 144 signal (replisome marker) were recorded for untreated cells. Directly after t = 0, ciprofloxacin 145 was introduced to the flow cell and a time-lapse was recorded over a period of 3 h. The numbers 146 of DinB-YPet foci per cell, reflective of pol IV binding events, were determined at 0, 30, 60, 90 147 and 120 min time points (Fig 1). Colocalisation between DinB-YPet foci and replisome was also  151 We first monitored pol IV behaviour in cells expressing wildtype levels of UmuD and 152 UmuC (EAW643 , Table 1). Cells exhibited no pol IV foci prior to ciprofloxacin treatment (Fig   153   1A, upper panel), as seen previously [12]. After ciprofloxacin addition, the number of pol IV foci 154 per cell increased to an average of 0.1 foci per cell from 60 min, i.e. one in ten cells exhibited a 8 155 pol IV focus. Consistent with our previous observations [12], the percentage of pol IV foci that 156 colocalised with the replisome dropped markedly between the 90 min and 120 min time-points 157 (Fig 1A, middle panel). From 0-90 min after ciprofloxacin addition, 5% of pol IV foci 158 colocalised with replisomes, somewhat less than observed previously. From 120-150 min this 159 decreased to < 2%. The percentage of replisome foci that contained a pol IV focus followed a 160 similar trend (Fig 1A, lower panel); from 0-90 min after ciprofloxacin addition, 0.5% of 161 replisome foci contained a pol IV focus, dropping to ~0.1% (indistinguishable from chance 162 colocalisation) from 120-150 min. 163 We next examined the effect of deleting the umuDC operon (and thus eliminating UmuD) 164 on the number of pol IV foci and the extent of colocalisation with replisomes (SSH007, Table   165 1). From 30 min, 10-15% of pol IV foci colocalised with replisomes. Compared to umuDC + 166 cells, ΔumuDC cells exhibited a 3-fold increase in the number pol IV foci per cell with ~0.3 foci 167 per cell from 60 min after ciprofloxacin addition (Fig 1B, upper panel). Moreover, deletion of 168 umuDC led to a 3-fold increase in the percentage of pol IV foci that colocalise with replisomes 169 (Fig 1B, middle panel). Interestingly, pol IV-replisome colocalisation now persisted above 10% 170 for the 90, 120 and 150 min time points. The percentage of replisomes that contained a pol IV 171 focus was also elevated in the ΔumuDC background (Fig 1B, lower panel). From 30 min, 2-4% 172 of replisomes contained a pol IV focus. Compared to umuDC + cells, this represents a 6-8-fold 173 increase in colocalisation. These observations suggest that the presence of UmuD (and/or 174 UmuD′) normally suppresses the formation of pol IV foci and also prevents pol IV from binding 175 within, or close to, replisomes. 176 We next explored the effects of expressing the non-cleavable UmuD(K97A) mutant in 177 dinB-YPet dnaX-mKate2 umuDC cells (SSH007 + pJM1243, Table 1). We note that the 9 178 observed effects might be slightly exaggerated because the expression level of UmuD(K97A) 179 from the plasmid is 4-5-fold higher than chromosomal levels [38,43]. At the 90 min time point, 180 cells contained on average 0.6 pol IV foci -a 6-fold increase over umuDC + cells (Fig 1C,   181 upper panel). This sixfold increase in pol IV foci per cell, was accompanied by a 3-fold increase 182 in colocalisation with replisomes (Fig 1C,  times higher expression levels than chromosomal expressed UmuD. These cells produced ~0.1 192 DinB-YPet foci per cell at 60 min (Fig 1D, upper panel), similar to umuDC + cells. In the cells 193 expressing UmuD′, colocalisation of pol IV with replisomes was generally low, but highly 194 variable (Fig 1D, middle panel). Two large spikes in colocalisation were apparent at the 30 and 195 90 min time points. However, due to the low number of foci available for analysis at these time-196 points, there was very large error associated with these values. No spikes in colocalisation were 197 observed when measuring the proportion of replisomes that contained a pol IV focus (Fig 1D,   198 lower panel). Importantly, the colocalisation of pol IV with replisomes decreased to < 1% after 199 90 min (Fig 1D, middle panel). Similarly, the percentage of replisomes with that contained a pol 200 IV focus drops between the 90 and 120 min time points (Fig 1D, lower panel). From 30-90 min, 10 201~1% of replisomes contained a pol IV focus. By 120 min < 0.1% of replisomes contained a pol 202 IV focus. Overall, the introduction of UmuD′ into umuDC cells restores rates of focus 203 formation and colocalisation with replisomes to near wild-type (umuDC + ) levels.  DinB-YPet images the foci formed in different strains appeared to exhibit differences in both 219 intensity and shape (Fig 2A-D, first row; 300 ms exposures). For the umuDC + (Fig 2A), 220 ΔumuDC (Fig 2B) and UmuD′-expressing cells (Fig 2D), most foci were relatively faint and 221 diffuse. In contrast, cells expressing UmuD(K97A) produced brighter, and much more distinct, 222 pol IV foci. Reasoning that these differences might reflect differences in the nature of pol IV 223 interactions with the substrates, we next measured the binding lifetime of pol IV at these sites.    For both umuDC + and umuDC cells, most pol IV binding at replisomes is short-lived 242 (Fig 2A, B). In the early stages of ciprofloxacin exposure (25-45 min) the components of the 243 autocorrelation function were 80% short-lived, 10% medium and 10% long-lived. In the later 244 stages, (120-150 min), the proportion of medium-long lived events increased to 40%. In cells 245 expressing UmuD(K97A) long-lived events were much more common: by the 120-150 min 246 period medium and long-lived events comprised 80% of the autocorrelation function (Fig 2C). In 12 247 stark contrast, cells expressing UmuD′ produced almost exclusively short-lived events (Fig 2D). 248 UmuD′ appeared to supress the medium and long-lived pol IV binding events that occur in wild-249 type umuDC + background following ciprofloxacin treatment.  Following ciprofloxacin treatment, cells typically contained multiple RecA* structures 270 (Fig 3A), including both foci at early time points, and more elongated "bundle" structures 271 described previously at later time points [38]. We next determined the percentage of DinB-YPet 272 foci that colocalised with RecA* features (Fig 3B). Prior to the introduction of ciprofloxacin, 273 RecA* filaments were rarely formed in cells during normal metabolism (< 0.1 mCI foci per cell) 274 consistent with our previous study [38]. Unsurprisingly, we did not detect colocalisation of pol damage, we expect exposed ssDNA substrates for RecA(E38K) binding to occur infrequently.
293 Therefore, we additionally tested whether constitutive SOS signalling may occur due to 294 constitutive RecA(E38K)-dsDNA filament formation. To that end, we tested the ability of 295 RecA(E38K) to form filaments on a 60-mer dsDNA substrate. We found that RecA(E38K) binds   produced few pol IV foci (Fig 4A) [12]. Cells that carried the SOS-constitutive lexA(Def) allele and the wild-type recA allele produced a relatively high level of DinB-YPet signal, but produced 316 few foci (Fig 4B) [12]. This result is consistent with our previous study in which we concluded 317 that binding is triggered by the presence of damage on the DNA, as opposed to mass action-318 driven exchange brought on by increased intracellular concentrations of pol IV [12]. In contrast 319 to both recA + strains, cells carrying both the lexA(Def) allele and the RecA*-constitutive 320 recA(E38K) allele produced both high DinB-YPet signal and readily visible foci (Fig 4C). These assemblies are themselves substrates for pol IV in these cells. 325 We therefore directly tested whether RecA(E38K) interacts with pol IV on filaments 326 assembled dsDNA in vitro. Using an identical SPR experimental setup [38], we assembled  1-10s, Fig 4F), indicative of pol IV binding to its target for longer periods (seconds 340 timescale).

341
To comprehensively assess pol IV binding lifetimes across all intensity trajectories, mean replisomes; 31% of replisomes experienced a pol IV binding event (Fig 4G, right panel). The 355 amplitude of the autocorrelation function was also increased (0.4 at Δt = 1 frame, Fig 4G, black   356 line), indicating that long-lived binding events occurred at replisomes much more frequently.

357
The decay rate of the autocorrelation function had two components (Fig 4G,  background. When analysing the binding behaviour of pol IV away from replisomes in these 361 three backgrounds, similar results were obtained (Fig 4H).

363
In this study, we arrived at four conclusions: i) UmuD promotes the binding of pol IV to       Peak fractions were identified as above and pooled. A DEAE-Sepharose column was not used.

529
Protein in this pool was precipitated by the addition of equal volume of 90% saturated 530 ammonium sulfate. The precipitate was stirred and then spun down at 13,000 rpm for 30 min.

531
The pellet was resuspended in R buffer plus 1 M ammonium sulfate, stirred for an hour, and then The surface was regenerated as previously reported [38]. Furthermore, the SPR signal 580 were corrected using a flow cell without immobilised bio-(dT) 71 or dsDNA and corrected for the 581 amount of immobilised RecA(E38K) [38]. Ghodke et al. utilised  Varian Cary 300 dual beam spectrophotometer equipped with a temperature controller and a 12-595 position cell changer. The cell path length and band pass were 0.5 cm and 2 nm, respectively.

596
The NADH extinction coefficient at 380 nm of 1.21 mM -1 cm -1 was used to calculate the rate of 597 ATP hydrolysis.

604
The nicked cdsDNA, buffer, and ATP regeneration system were preincubated at 37 ˚C for 10   extracted from the fit error using the two-exponential fit (Suppl. Fig 1G, H). The error bar from