The E. coli strains used in this study are listed in Table 1. The λRED system was used to construct chromosomal mutations (Datsenko and Wanner, 2000). Briefly, PCR fragments including mutated gly-V-X-Y operon or datA and frt-kan gene was transformed into the strain bearing pKD46 (λ RED expression plasmid). To verify the presence of specific mutations in each strain constructed in this study, the chromosomal DNA regions of interest were PCR amplified and analyzed by Sanger sequencing. The detail of λRED system was described below. All bacterial strains were grown in LB medium or M9 medium supplemented with 0.2% casamino acids, 5 μg/mL thiamine, and 0.2% glucose.

A datA fragment spanning P_orn to P_queG was amplified by PCR from MG1655 genomic DNA and primers datA_LI and datA_RI (see Table 2 for primer sequences), digested with AatII and HindIII-HF, and inserted into the pACYC177 to yield pITK1. pITK1-derivatives for construction of genomic mutants were generated using inverse PCR with the following primers (Table 2): ΔPorn_L and ΔPorn_R for pITK2 (ΔP_orn), ΔPgly1_L and ΔPgly1_R for pITK3 (ΔP_glyV-X-Y), delter_L and delteronly_R for pITK12 (ΔT_glyV-X-Y), and subDnaAbox2_L and subDnaAbox2_R for pITK15 (datAsubDnaAbox2). A glyV-X-Y fragment was amplified by PCR from MG1655 genomic DNA using primers datAess_L and datAess_R, digested with BglII and HindIII-HF, and inserted into the pKP1673 to yield pITK6. A frt-kan fragment was amplified by PCR using pTH5 and primers 40kan-T1T2-L and 40kan-T1T2-R, digested with EcoRI-HF, and ligated with a EcoRI-HF-digested vector fragment that was amplified by inverse PCR using pBAD18 and primers 18-T1T2-L and 18-T1T2-R to yield pSR14 carrying frt-kan instead of the bla gene.

For the construction of mutant strains in Figures 3, 5, 6, datA fragments carrying the desired mutation were amplified from mutant plasmids (pITK2 for ΔP_orn, pITK3 for ΔP_glyV-X-Y, pITK12 for ΔT_glyV-X-Y and pITK15 for datA subDnaAbox2) with primers datAfusion_L2 and datA_R_f (Table 2). Using the overlap extension (OE)-PCR method (Hilgarth and Lanigan, 2020), these fragments were joined with a frt-kan fragment amplified from pTH5 using primers pTH5-kan_L_f and pTH5-kan_R. Fragments were transformed into MG1655 or SH022 cells bearing pKD46 by electroporation, and the frt-kan was removed by introduction of the pCP20 plasmid, yielding strains ITK4 (ΔP_orn), ITK5 (ΔP_glyV-X-Y), ITK8 (ΔT_glyV-X-Y), ITK16 (datAsubDnaAbox2), ITK4c (dnaC2ΔP_orn), ITK5c (dnaC2ΔP_glyV-X-Y), ITK8c (dnaC2ΔT_glyV-X-Y), and ITK16c (dnaC2 datAsubDnaAbox2). For generation of the ITK3 control strain, a frt-kan fragment amplified from the pTH5 plasmid with primers pTH5-kan_L_f and pTH5-kan_R was transformed into MG1655 cells bearing pKD46. For the construction of mutant strains in Figure 4, fragments carrying a transcription terminator sequence of the rrnB operon (rrnBT) were amplified from pSR14 using primers T1T2-frtkan-L and T1T2-frtkan-R for rrnBT-U, and T1T2-queGP-L and T1T2-frtkan-R for rrnBT-D, and were transformed into MG1655 or SH022 cells bearing pKD46 by electroporation. The frt-kan was removed by introduction of the pCP20 plasmid, yielding strains SR21 (rrnBT-U), SR22 (rrnBT-D), and SR33 (dnaC2 rrnBT-U). A fragment carrying frt-kan was amplified from pTH5 with primers PqueG-U-3 and PgueG-del-L2, and was transformed into MG1655 or SH022 cells bearing pKD46 by electroporation. The frt-kan was removed by introduction of the pCP20 plasmid, yielding strains SR18 (ΔP_queG) and SR34 (dnaC2 ΔP_queG). A fragment carrying the datAsubIBS mutation and frt-kan was amplified from pKX40kan (Kasho et al., 2017) with primers ChdatA-U and ChdatAsubIBS, and transformed into MG1655 or SH022 cells bearing pKD46 using electroporation. The frt-kan was removed by introduction of the pCP20 plasmid, yielding strains ITK24 (datAsubIBS) and ITK24c (dnaC2 datAsubIBS), respectively. For checking the introduction of specific mutations in each strains constructed in this study, the sequences of chromosomal datA locus were confirmed by DNA sequencing.

Chromatin affinity precipitation (ChAP, a modified version of ChIP assay) and qPCR experiments were performed according to a previously described method (Kasho et al., 2014, 2021; Inoue et al., 2016). In brief, SH022 (dnaC2 ihfA-cHis12) or its derivatives were grown in LB medium at 30°C, a permissive temperature, until A₆₆₀ of 0.04 was reached, after which cells were incubated at 37°C, a restrictive temperature, for 80 min (“0 min” sample). Cells were then incubated at 30°C for 5 min to initiate replication (“5 min” sample), followed by further incubation at 38°C for 5–25 min to allow a single round replication to proceed (“10–30 min” samples). For preparing random culture (“R”) samples, SH022 cells or their derivatives were grown in LB medium at 30°C until the A₆₆₀ reached 0.15. Samples were cross-linked with 3% (final) formaldehyde at room temperature for 5 min. The reactions were quenched by incubation in 125 mM glycine for an additional 5 min. Then, the cells were collected by centrifugation, washed twice with 1 mL of ice-cold TBS [50 mM Tris–HCl (pH7.5) and 500 mM NaCl], resuspended in 500 μL of binding buffer [50 mM Tris–HCl (pH 7.5), 500 mM NaCl, 1% (vol/vol) Triton X-100, 5 mM imidazole, and 0.1 mM PMSF], and sonicated six times for 20 s each. The resulting size of the chromosomal DNA was about 1 kb. Cell debris was then removed by centrifugation at 16,000 × g for 15 min at 4 °C, and a portion (400 μL) of the resulting supernatant (450 μL) was mixed with 10 μL of Dynabeads His-tag Isolation and Pulldown (lifetechnologies), followed by incubation at 4°C for 1 h with a gentle rotation. Beads and bound materials were washed four times with wash buffer [50 mM Tris–HCl (pH 7.5), 500 mM NaCl, 1% (vol/vol) Triton X-100, and 5 mM imidazole], resuspended in elution buffer [50 mM Tris–HCl (pH 7.5), 500 mM imidazole, 1% SDS, 10 mM EDTA, and 10 mM dithiothreitol], and incubated at 65°C for 12 h to allow de-crosslinking. DNA in the samples before (Input) and after (ChAP) pull down was purified using a Wizard SV Gel and PCR Clean-Up System (Promega). The levels of oriC, datA, and ylcC were quantified by real-time qPCR using TB Green Premix Ex TaqII (Tli RNaseH Plus) (TaKaRa) and primers ORI_1 and KWoriCRev for oriC; RT-datAIBS-L and RT-datAIBS-R for datA; and RTYLCC-L and RTYLCC-R for ylcC.

Total cellular RNA samples used in Figures 2B, 3B, 5, and Supplementary Figure 1 were purified using NucleoSpin™ RNA (Macherey-Nagel) according to the manufacturer’s protocol. Total cellular RNA samples used in Figures 3E–I, 4, 6, 7 were purified using the hot phenol method (Su’etsugu et al., 2003). In brief, KYA018 (dnaC2), SH022, or their derivative cells were grown in LB medium at 30°C until they reached A₆₆₀ of 0.04, followed by further incubation at 37°C for 80 min. Cells were then incubated at 30°C for 5 min, followed by further incubation at 38°C for 5–25 min. Samples were withdrawn at each of the indicated time-points and immediately mixed with equal volume of Stop buffer (20 mM sodium acetate, pH 5.2, 2% 30 mM sodium acetate-saturated phenol, pH 5.0, 2 mM EDTA, and 75% ethanol in diethylpyrocarbonate-treated MilliQ) to inactivate cellular RNase activity. The cells were then collected by centrifugation and resuspended with 400 μL of Resuspension buffer (20 mM sodium acetate, pH 5.2, 2% sodium dodecyl sulfate in diethylpyrocarbonate-treated MilliQ). Total RNA was then extracted twice by treatment with 400 μL 30 mM sodium acetate-saturated phenol (pH 5.0) and three times with 400 μL phenol/chloroform/isoamyl alcohol (25:24:1), collected by ethanol precipitation, and suspended with 50 μL RNase free dH₂O (TaKaRa). The concentration of purified total RNA samples was determined using a NanoDrop™ Lite (Thermo Scientific).

For RT-qPCR, cDNA samples were prepared using 1 μg of each total RNA sample and PrimeScript™ RT reagent Kit with gDNA Eraser (Perfect Real Time) (TaKaRa). The RNA levels were quantified by real-time qPCR using TB Green Premix Ex Taq™II (Tli RNaseH Plus) (TaKaRa) and primers ORI_1 and KWoriCRev for oriC; RT-datAIBS_L and RT-datAIBS_R for datA; RTrpoA-L and RTrpoA-R for rpoA; and RTmioC-L and RTmioC-R for mioC.

Northern blotting experiments were performed according to a previously described method (Su’etsugu et al., 2003), with the following minor modifications. Each transcript in purified total RNA samples was separated by electrophoresis using 2% agarose/2.2 M formaldehyde gel in 1x MOPS running buffer, transferred to Hybond™ N+ membrane (Cytiva) using the capillary blotting method, dried at 80°C for 2 h, and then fixed by exposure to UV (120 mJ/cm²) for 5 min using FUNA-UV-Crosslinker FS-800 (Funakoshi). Oligonucleotide probes with complementary sequences to either the datA IBS locus or the rpoA gene were end-labeled with ³²P, and this labeled probe was hybridized with RNA on the membrane in the presence of ULTRAhyb™ Ultrasensitive Hybridization Buffer (Invitrogen). The membrane was then washed with 2x SSC (300 mM sodium chloride, 30 mM sodium citrate), 2x SSC+0.1% sodium dodecyl sulfate, 0.2x SSC (30 mM sodium chloride, 3 mM sodium citrate) +0.1% sodium dodecyl sulfate, and 0.2x SSC, dried, and analyzed using the Typhoon™ 7500 imaging analyzer (Cytiva).

The minimal region of datA required for full DDAH activity is located at the intergenic region downstream of the tRNA-Gly (glyV-X-Y) operon between the yjeV gene and queG gene (Figure 2A). Indeed, the datA IBS was transcribed in exponentially-growing MG1655 (WT) cells when cultivated at 37°C in LB or supplemented M9 medium (Supplementary Figure 1A). Thus, we hypothesized that datA-IHF binding/dissociation is regulated by transcription of the datA locus. This theory is in accordance with previous observations that rifampicin inhibits timely dissociation of datA-IHF complexes (Kasho and Katayama, 2013), and that datA activity is inhibited by the translocation of datA to a highly transcribed position (Frimodt-Møller et al., 2015).

To analyze the cell cycle-dependent correlation between datA-IHF binding and transcription of the datA locus, we combined RT-qPCR and ChAP assays using temperature sensitive dnaC2 cells, which allow synchronization of the cell cycle in LB medium (Withers and Bernander, 1998). DnaC is the helicase loader, and the dnaC2 mutation inhibits replication initiation at oriC specifically at restrictive temperatures (37–42°C). Thus, the cell cycle of the dnaC2 cells can be synchronized by incubation at 37°C for 80 min just before replication initiation (0 min), followed by a temperature down shift to 30°C for 5 min to release the cell cycle by activating DnaC, and induce a single round of replication initiation. An upshift back to 37°C for 10–30 min allows DNA synthesis to proceed while inhibiting additional replication initiation. This dnaC2-based synchronization system has previously been used to identify the cell cycle-coordinated IHF binding/dissociation pattern at oriC, datA, and DARS2 (Kasho and Katayama, 2013; Kasho et al., 2014), as well as the cell cycle-specific oscillation of expression levels of genes including dnaA, mioC, and gidA (Theisen et al., 1993; Su’etsugu et al., 2003); thus, this well-established method is suitable for the purpose of this study.

To validate our assay, initial control experiments were performed in which the cell cycle-coordinated binding/dissociation of IHF at the datA and oriC loci was confirmed using ChAP-qPCR. In accordance with previous studies (Kasho and Katayama, 2013), at the oriC locus IHF was stably bound prior to initiation (0 min) and immediately dissociated after initiation (5–10 min) (Supplementary Figure 1B), while at the datA locus, IHF was not bound before initiation (0 min) and became bound after initiation (15–20 min) (Figure 2B).

Next, under the same experimental conditions as in Figure 2B, we prepared total RNA samples and performed RT-qPCR using primers recognizing datA IBS, the mioC gene as a control for cell cycle synchronization, or the housekeeping gene rpoA gene as a loading control (Su’etsugu et al., 2003). Consistent with a previous report, the mRNA level of the mioC gene relative to that of rpoA was lower at 0 min (Supplementary Figure 1D), indicating that the dnaC2 cells were properly synchronized before initiation. By contrast, the relative RNA level at datA IBS was very low in the non-synchronized cells, dramatically increased at the initiation period (0–5 min), and decreased after initiation (10–20 min) (Figure 2B; Supplementary Figure 1C). No dramatic increase in the datA RNA level or decrease in the mioC mRNA level was observed when MG1655 (dnaC WT) cells were incubated at 37°C (Supplementary Figures 1C, D), indicating that transcription of datA is regulated in a cell cycle-dependent manner. A similar oscillation pattern of datA RNA level was observed using supplemented M9 medium (Supplementary Figure 1E). Thus, these results indicated that transcription of datA is regulated in a cell cycle-dependent manner.

In dnaC2 cells, the level of datA RNA was approximately 8–40-fold higher at initiation (0 min) than in the random culture (R) sample (Figure 2B; Supplementary Figure 1C). The datA RNA levels were still higher at the post-initiation period in dnaC2 cells than in random culture samples, which can also be explained by the accumulation of datA RNA as a result of cell cycle synchronization. Furthermore, the accumulation of datA RNA might affect the overall high baseline datA RNA level after releasing the cell cycle, i.e., during 5–10 min incubation, it is possible that datA RNA is not significantly degraded but is also newly produced. Thus, we concluded that the levels of IHF binding and transcription at the datA locus are inversely correlated with each other, supporting the idea that the datA-IHF binding/dissociation is regulated in a manner dependent on transcription.

To determine the transcriptional start site of datA RNA, we performed RT-qPCR using four primer sets for the orn gene, two sets each for the region upstream or downstream of the transcriptional promoter for glyV-X-Y operon (P_glyV-X-Y), and one for datA IBS (Figure 3A). While the RNA levels of the orn gene and the glyV-X-Y transcription units remained constant, the relative RNA level of datA IBS fluctuated during the cell cycle and significantly increased at initiation, consistent with Supplementary Figure 1C (Figure 3B; probe 9). The relative mRNA levels of orn gene (probes 1–4) and the region upstream of the transcriptional promoter P_glyV-X-Y (probes 5 and 6) were much lower than those of the glyV-X-Y operon (probes 7 and 8) or datA IBS (probe 9), suggesting that transcription of the orn gene is substantially terminated at the P_glyV-X-Y upstream locus.

To confirm whether datA transcription is initiated at P_glyV-X-Y, northern blotting was used to analyze the length of the datA-containing RNA products. The relative amounts of datA-IBS RNA were low (Figure 3B), which may result from degradation of the 3′-region of RNA products including tRNA-Gly during the tRNA maturation processes (Wellner et al., 2018). Thus, to efficiently analyze full-length RNA products that read through the datA IBS locus, we constructed the ΔglyV-X-Y::kan strain, in which the entire glyV-X-Y operon was replaced by the kan gene (Figures 3C, D). To avoid cell lethality resulting from lack of tRNA-Gly, tRNA-Gly was supplemented from the pITK6 plasmid-encoded glyV-X-Y in all the ΔglyV-X-Y::kan mutant cells constructed in this study.

In the ITK9c (dnaC2 ΔglyV-X-Y::kan) cells, the level of datA RNA was significantly lower than that of the kan gene (Figures 3B, E; “R” sample), suggesting that transcription inhibition occurs at the datA-upstream region. Furthermore, the datA RNA level was 3-fold higher at initiation compared with randomly-growing cells (Figure 3E; “0 min” sample). Consistent with the RT-qPCR experiment, the northern blotting experiments revealed that the read-through RNA products, including datA IBS (approximately 1,500 nt), were observed using either the datA IBS probe or the kan probe (Figures 3F, G) and were increased at the initiation period (Figures 3H, I). Notably, the shorter RNA intermediate (approximately 800 nt) was only observed using the kan probe (Figure 3F). Given the observed fluctuation of datA-IBS RNA, these data suggest that the transcripts initiated from P_glyV-X-Y pass through datA IBS specifically at the initiation period and are terminated in front of datA IBS after initiation. We thus hypothesized that the specific DNA element located between the region downstream of glyV-X-Y and datA IBS is required for the cell cycle-coordinated datA-IBS transcription. Consistently, the relative RNA levels initiated from P_glyV-X-Y were not increased at the initiation period in either WT or ΔglyV-X-Y::kan cells (Figures 3B, E), supporting our hypothesis. The relative datA-IBS transcription levels in Figure 3B were moderately lower than those in Figure 2B, which might be due to subtle differences in cell-cultivation conditions affecting rates of transcription and processing of the full-length datA RNA products, including tRNA-Gly.

Next, to address the requirement of P_glyV-X-Y for transcription-dependent regulation of datA-IHF binding, we constructed ITK4c (dnaC2 ΔP_orn) and ITK5c (dnaC2 ΔP_glyV-X-Y) strains by inserting frt-kan into the nnr gene located close to, but outside of, the datA region (Figure 4A). In random cultures (Figures 4B–E), the nnr::frt-kan mutation had little effect on cell cycle-coordinated datA-IHF binding and datA transcription (Figures 4B, C), or oriC-IHF binding and mioC mRNA level (Figures 4D, E). We thus used ITK3c (dnaC2 nnr::frt-kan) cells as a WT control for analyzing cell cycle-coordinated oscillation of datA RNA and datA-IHF binding/dissociation.

In cells synchronized at the initiation period (Figures 4F, G), introduction of the ΔP_glyV-X-Y mutation greatly decreased the datA-IBS RNA level compared with that of ΔP_orn and the WT control and, notably, diminished the initiation period-specific decrease in datA-IHF binding (Figures 4F, G). Low residual RNA levels could be due to read-through RNA products from P_orn. In contrast to datA IBS RNA, the mioC mRNA level was not substantially changed by introduction of the ΔP_glyV-X-Y mutation (Figure 4H). There may be a minor increase in the oriC-IHF binding level in ΔP_glyV-X-Y mutant cells, which could be an indirect consequence of a moderate delay in replication initiation induced by increased datA-IHF binding (see below). Consistent with qPCR analysis, northern blotting supports the requirement of P_glyV-X-Y for the initiation period-specific production of read-through datA RNA (Figures 4J, K).

To address the impact of the ΔP_glyV-X-Y mutation on regulation of replication initiation, we performed flow cytometry analysis. Cells were exponentially grown in LB medium and further incubated in the presence of rifampicin and cephalexin, which inhibit replication initiation and cell division, respectively, to complete replication. Cell mass (cell size) and chromosomal DNA content were then analyzed by flow cytometry to allow determination of the oriC/cell and the ori/mass, and thus the frequency of replication initiation. While WT cells predominantly showed peaks corresponding to four and eight origins, ΔP_glyV-X-Y, but not ΔP_orn, cells showed a significant increase in the peak corresponding to four origins and a decrease in the eight origins peak (Figure 4L). This means that the ΔP_glyV-X-Y mutation specifically decreased initiation frequency (ori/cell or ori/mass) (Figure 4L), consistent with our observation that P_glyV-X-Y is required for production of datA RNA and transcription-dependent regulation of datA-IHF binding. Given that IHF dissociates from oriC immediately after initiation, delay of initiation could result in an increased level of oriC-IHF complexes (Figure 4I).

To confirm the requirement of P_glyV-X-Y for regulation of datA function, we inserted the frt site-flanking strong transcriptional terminator element (rrnBT; terminator of E. coli rrnB operon) (Orosz et al., 1991) either upstream or downstream of datA IBS (rrnBT-U or -D, respectively) using the inserting frt site (Figure 5A). The rrnBT-U cells, but not the rrnBT-D or ΔP_queG cells, had significantly decreased levels of datA-IBS RNA compared with WT cells (Figure 5B). Consistent results were obtained by Northern blotting where the read-through RNA products including datA IBS (500–1,000 nt) were observed in WT, rrnBT-D, and ΔP_queG cells, but not in rrnBT-U cells (Figures 5C, D). Importantly, insertion of rrnBT-U specifically impaired timely datA-IHF dissociation and datA transcription at the initiation period, while stable oriC-IHF binding was maintained in rrnBT-U cells (Figures 5E–L). An observed slight increase in levels of oriC-IHF binding in rrnBT-U cells could be due to delayed initiation as for ΔP_glyV-X-Y (Figure 5L). Consistently, rrnBT-U cells showed moderately delayed initiation frequency in WT, ΔDARS1, or ΔDARS2 backgrounds, whereas ΔP_queG cells did not (Figure 5G; Supplementary Figure 3). Also, in the rrnBT-U cells, rrnBT was inserted at the region upstream of T_glyV-X-Y to avoid affecting P_yjeV transcription, thus suggesting that P_glyV-X-Y has a predominant role in transcription-dependent datA-IHF regulation and the requirement of P_yjeV is limited. Taken together, these observations support the idea that P_glyV-X-Y is critical for the transcriptional regulation of timely datA-IHF binding.

To address the regulatory mechanism and the requirement of the specific DNA element located between the region downstream of the glyV-X-Y operon and the datA IBS for transcription-dependent datA regulation, we used substitution mutants at datA DnaA box2 and IBS (subDnaAbox2 and subIBS, respectively) and a deletion mutant in which a predicted terminator/attenuator sequence T_glyV-X-Y of the glyV-X-Y operon was removed (ΔT_glyV-X-Y) (Figure 6A; Ju et al., 2019). Because a previous study suggested that IHF binding at the promoter-attenuator region of the ilvG-M-E-D-A operon induces the pausing of transcription machinery (Pagel and Hatfield, 1991), we first tested whether datA-IHF binding interrupts transcription. For this purpose, we analyzed transcription levels of the datA IBS region in dnaC2 or dnaC2 datAsubIBS double mutant cells using RT-qPCR and primer set No. 9 from Figure 3A because these primers hybridize to sites outside of the datA IBS.

In dnaC2 datA WT cells, datA RNA was abundant during the initiation period (0–5 min) and decreased after initiation (20–30 min) (Figures 6B, C). By contrast, introduction of the datAsubIBS mutation diminished this oscillation and the level of datA RNA remained constant. The constant increase of datA RNA throughout cell cycle resulted in accumulation of transcript and, therefore, an overall increase in datA RNA levels relative to WT cells in random culture samples (Figure 6B). Consistent with Figures 6B, C, introduction of the datAsubIBS mutation into dnaC2 ΔglyV-X-Y::kan cells resulted in increased total datA RNA levels (Figure 6D), and notably also increased the level of read-through datA RNA (datA IBS or kan probes; approximately 1,500 nt) relative to that of the RNA intermediate (kan probe; approximately 900 nt) (Figures 6D–F). This suggests that datA-IHF binding at the post-initiation period is stable enough to significantly interrupt transcription. However, a slight increase in the datA RNA level at the initiation period was observed, even in the dnaC2 datAsubIBS cells (Figures 6B, C), implying that other potential regulatory elements may exist to interrupt datA transcription.

To address the possible role of T_glyV-X-Y for regulating transcription of the datA locus, we constructed a deletion mutant of T_glyV-X-Y (ΔT_glyV-X-Y; Figure 6A) (Ju et al., 2019). T_glyV-X-Y overlaps with datA DnaA box 1 and the −35 sequence of a putative promoter, P_yjeV (Kitagawa et al., 1996). Therefore, we chose to delete datA DnaA box 1, which is dispensable for DDAH activity (Kasho and Katayama, 2013; Kasho et al., 2017), but not P_yjeV. In addition, because previous studies suggested that DnaA binding around the terminator sequence of the asnC gene inhibits the passage of transcription machinery (Schaefer and Messer, 1988, 1989), we also addressed the possibility that the DnaA binding at the upstream region of datA IBS can act as a roadblock for transcription from P_glyV-X-Y. Since datA DnaA box 2 is essential for ATP-DnaA-specific oligomer formation at datA, we constructed a datA mutant with DnaA box 2 substitution (datAsubDnaAbox2) (Figure 6A), which lacks the DDAH activity of datA (Kasho and Katayama, 2013; Kasho et al., 2017).

Introduction of ΔT_glyV-X-Y or the datAsubDnaAbox2 mutation resulted in increased levels of datA-IBS RNA at the initiation period (0–5 min) (Figures 7A, B). In contrast to the datAsubIBS mutant cells (Figures 6B, C), both the ΔT_glyV-X-Y and datAsubDnaAbox2 mutant cells sustained cell cycle-coordinated oscillation of datA RNA levels, although datA RNA levels at the post-initiation period (20–30 min) were also increased compared with those of WT cells. This suggests that either T_glyV-X-Y or datA DnaA box 2 may assist in regulation of the overall transcription level of datA but are not essential for the timely dissociation of the datA-IHF complex. Consistently, northern blotting analysis of datA-containing transcripts showed that the level of read-through datA RNA (IBS or kan probes; approximately 1,500 nt) relative to the intermediate (kan probe; approximately 900 nt) was only moderately increased in ΔT_glyV-X-Y and datAsubDnaAbox2 cells compared with WT cells (Figures 7C–E). Despite the increase in basal datA transcription levels, the ΔT_glyV-X-Y mutation had little effect on initiation timing (Figure 7F). These results suggest that transcription block by T_glyV-X-Y or the DnaA binding at datA DnaA boxes 2 play specific roles in regulating the basal level of datA transcription but is dispensable for its cell cycle-coordinated oscillation and for controlling the timing of initiation. Taken together, these results suggest that another regulatory factor may play a role in timely datA-IHF binding in a manner independent of datA transcription.