Mycobacterium smegmatis Lhr is a DNA-dependent ATPase and a 3'-to-5' DNA translocase and helicase that prefers to unwind 3'-tailed RNA:DNA hybrids

Background : Helicases play important roles in nucleic acid transactions. Results : Mycobacterial Lhr uses the energy of ATP hydrolysis to drive 3'-to-5' translocation along single strand DNA and to unwind duplexes en route . Conclusions : Lhr is better at unwinding an RNA:DNA hybrid than a DNA:DNA duplex. Significance : Lhr defines a novel clade of helicases with a signature domain organization.

DNA helicases are nucleic acid-dependent NTPases that play important roles in DNA replication, recombination and repair. RNA helicases orchestrate transcription, RNA processing, ribosome biogenesis, translation and RNA turnover. Helicases use the chemical energy of NTP hydrolysis to either effect mechanical changes in the secondary structure of nucleic acids or to remodel the structures of protein-nucleic acid complexes. Helicases have been classified into superfamilies, families, and subfamilies according to their distinctive primary, tertiary and quaternary structures and their biochemical specificities: i.e., their NTP preference, nucleic acid preference, and directionality of translocation or unwinding (1).
Differences in the roster of DNA helicases between taxa offer useful clues to the evolution and diversification of replication/repair strategies and, where the rosters diverge in animals versus infectious pathogens, they can suggest antiinfective drug targets. With this in mind, we are focused on the ensemble of DNA helicases in Mycobacteria, a genus of the phylum Actinobacteria that includes the human pathogen M. tuberculosis and its avirulent relative M. smegmatis. To date, seven mycobacterial DNA helicases have been purified and characterized: UvrD1, UvrD2, and AdnAB are superfamily I (SF1) helicases; XBP, RqlH, SftH and RecG are superfamily II (SF2) helicases. UvrD1 (2-6), UvrD2 (2,7,8), XBP (9)(10)(11), RqlH (12), and SftH (13) are monomeric proteins that unwind 3'-tailed duplex DNAs by loading onto the 3' single strand and translocating in the 3'-to-5' direction. RecG unwinds multi-strand DNA junctions that resemble stalled replication forks or recombination intermediates (14). AdnAB is a heterodimeric motor-nuclease machine that unwinds and resects DNA double strand breaks during homologous recombination (15)(16)(17)(18). The 3'-to-5' translocasehelicase activity of the AdnB subunit drives unwinding from a duplex end, during which the AdnA and AdnB nucleases incise the displaced 5' and 3' strands, respectively. Of the seven biochemically characterized helicases, only UvrD2 has been shown to be essential for mycobacterial growth in culture (7,8).
Here, we identify, purify and biochemically characterize a new mycobacterial SF2 helicase named Lhr, in deference to its extensive homology to the Lhr polypeptide of Escherichia coli. The E. coli lhr gene ("long helicase-related") was originally identified and studied by the Deutscher lab because of its location immediately downstream of the rnt gene encoding the tRNAprocessing enzyme RNase T (19). The 1538-aa Lhr polypeptide was, at the time, the longest known E. coli protein; it is presently the fifth largest E. coli protein after YeeJ (2358-aa), YdbA' (2020-aa), YneoO' (1806-aa) and YfhM (1653-aa). Although E. coli Lhr was predicted to be an SF2 helicase, Reuven et al. (19) were unable to demonstrate any nucleic-acid-dependent ATPase activity associated with this protein. Disruption of the E. coli lhr gene by insertion of a kanamycin-resistance cassette near the 5' end of the ORF (which interrupts the putative motor domain) had no effect on E. coli growth under a wide variety of conditions tested (19). The lhr::kan mutation did not sensitize E. coli to killing by UV or peroxide and there was no synthetic growth phenotype when lhr::kan was combined with recA, recB, recD, uvrD, uvrB, uvrD or rep (19).
The 1507-aa M. smegmatis Lhr protein and the corresponding 1513-aa M. tuberculosis Lhr protein are encoded by genes Msmeg_1757 and Rv3296, respectively.
The N-terminal third of mycobacterial Lhr resembles the NTPase domains of SF2 helicases, as defined by the following characteristic motifs (highlighted in Fig. 1): (i) motif I GxGKT (the P-loop or Walker A-box), the lysine side chain of which contacts the β and γ phosphates of the NTP substrate and the serine of which coordinates the metal cofactor; (ii) motif II DExH (the Walker B-box), which coordinates the metal cofactor; (iii) motif VI QxxGRxGH, which coordinates the NTP γ phosphate and the water nucleophile for NTP hydrolysis; and (iv) motif III SAT, which makes bridging contacts to motifs II and VI and couples NTP hydrolysis to motor activity (20,21). The C-terminal two-thirds of Lhr has no instructive motifs or homologies to any proteins outside the clade of Lhr homologs found in many diverse bacterial taxa (Table S1). To the extent that the genetic neighbor might suggest biological function, it is worth noting that the nei gene immediately downstream of lhr in M. smegmatis and M. tuberculosis encodes the DNA repair enzyme endonuclease VIII (22).
To our knowledge, there has been no prior characterization of any bacterial Lhr enzyme. To fill this knowledge gap, we produced, purified and characterized recombinant M. smegmatis Lhr. Key questions of interest to us were: (i) Is Lhr a nucleic acid-dependent phosphohydrolase and, if so, what is its substrate and cofactor specificity? (ii) Can Lhr couple NTP hydrolysis to mechanical work, especially duplex unwinding? (iii) Do the biochemical activities of Lhr point to a role in DNA or RNA transactions? and (iv) Is the Cterminal domain of Lhr required for its activities?

EXPERIMENTAL PROCEDURES
Lhr proteins. The ORFs encoding full-length M. smegmatis Lhr (Msmeg_1757) and C-terminal truncations Lhr-(1-562), Lhr-(1-856), and Lhr-(1-1098) were PCR-amplified from M. smegmatis genomic DNA with primers that introduced a NdeI site at the start codon and a XhoI site downstream of the native or newly installed stop codons. The PCR products were digested with NdeI and XhoI and ligated into pET16b that had been digested with NdeI and XhoI. The resulting pET-Lhr plasmids encode His 10 -tagged Lhr polypeptides under the transcriptional control of a T7 RNA polymerase promoter. The plasmid inserts were sequenced to verify that no unintended coding changes were acquired during amplification and cloning. The pET16b-LHR plasmids were transformed into E. coli BL21(DE3) cells. Cultures (1-liter) amplified from single ampicillin-resistant transformants were grown at 37°C in LB broth containing 100 µg/ml ampicillin until the A 600 reached 0.5. The cultures were chilled on ice for 1 h, then adjusted to 2% (v/v) ethanol and 0.3 mM isopropyl-β-D-thiogalactopyranoside and incubated for 16 h at 17°C with constant shaking. All subsequent procedures were performed at 4°C. Cells were harvested by centrifugation and resuspended in 25 ml of buffer A (50 mM Tris-HCl, pH 8.0, 500 mM NaCl, 1 mM DTT, 10% sucrose) containing 1 protease inhibitor cocktail tablet (Roche). Lysozyme was added to a concentration of 1 mg/ml and the suspension was incubated for 30 min before adding Triton X-100 to a final concentration of 0.1%. After incubation for an additional 30 min, the lysate was sonicated to reduce viscosity and the insoluble material was pelleted by centrifugation at 38000g for 1 h. The supernatant was mixed for 1 h with 2 ml of Ni-NTA agarose resin (Qiagen) that had been equilibrated with buffer A. The resin was recovered by centrifugation and resuspended in 30 ml of buffer A. The resin was again recovered by centrifugation, then resuspended in 10 ml of buffer B (50 mM Tris-HCl, pH 8.0, 500 mM NaCl, 0.05% Triton X-100, 10% glycerol) containing 50 mM imidazole and poured into a column. After washing the column with 4 ml of 3 M KCl, the bound material was eluted stepwise with 4-ml aliquots of 50, 100, and 500 mM imidazole in buffer B. The polypeptide compositions of the eluate fractions were monitored by SDS-PAGE. The recombinant Lhr proteins were recovered in the 100 and 500 mM imidazole fractions, which were pooled and then dialyzed overnight against 4 liters of buffer C (50 mM Tris-HCl, pH 8.0, 500 mM NaCl, 1 mM DTT, 1 mM EDTA, 0.05% Triton X-100, 10% glycerol). In order to deplete residual nucleic acids, the dialyzed Lhr preparations were applied to 2-ml DEAE-Sephacel columns that had been equilibrated with buffer C and the Lhr proteins were recovered in the flow-through fraction. Total protein concentrations of the DEAE preparations were gauged with the BioRad dye reagent using BSA as the standard. The concentrations of the full-length Lhr and truncated Lhr-(1-562), Lhr-(1-856), and Lhr-(1-1098) polypeptides were then determined by SDS-PAGE analysis of aliquots of each preparation (5 µg) in parallel with 1.25, 2.5, 5 and 10 µg of a BSA standard. The gel was stained with Coomassie blue and the staining intensities of the Lhr and BSA polypeptides were quantified using a Chemidoc XRS imager and Quantity One software. Lhr concentrations were calculated by interpolating the staining intensities to the BSA standard curve. The yields of Lhr and Lhr truncations were ~25 mg per liter of culture.
Nucleoside triphosphatase assay. Reaction mixtures containing (per 10 µl) 20 mM Tris-HCl, pH 7.0, 1 mM DTT, 1 mM CaCl 2 , 1 mM [α 32 P]ATP (Perkin-Elmer Life Sciences), and DNA and Lhr as specified were incubated for 30 min at 37˚C. The reactions were quenched by adding 2 µl of 5 M formic acid. An aliquot (2 µl) of the mixture was applied to a polyethyleneiminecellulose TLC plate, which was developed with 0.45 M ammonium sulfate. The radiolabeled material was visualized by autoradiography and 32 P-ADP formation was quantified by scanning the TLC plate with a Fujix BAS2500 imager.
Streptavidin displacement assay of Lhr translocation on DNA. Synthetic 34-mer oligodeoxynucleotides of otherwise identical nucleobase sequence containing a Biotin-ON internucleotide spacer either at the fourth position from the 5' terminus or the second position from the 3' terminus were purchased from Eurofins MWG Operon. These DNAs were 5' end-labeled with [γ 32 P]ATP by using T4 polynucleotide kinase and then purified by electrophoresis through a native 18% polyacrylamide gel. Streptavidin-DNA (SA-DNA) complexes were formed by preincubating 100 nM biotinylated 32 P-DNA with 4 µM streptavidin (Sigma) in 20 mM Tris-HCl, pH 7.0, 1 mM DTT, 1 mM CaCl 2 , and 1 mM ATP for 15 min at room temperature. The mixtures were supplemented with 40 µM free biotin (Fisher) and the displacement reactions (10 µl, containing 1 pmol biotinylated 32 P-DNA) were initiated by adding 10 pmol Lhr. After incubation for 15 min at 37˚C, the reactions were quenched by adding 3 µl of a solution containing 200 mM EDTA, 0.6% SDS, 25% glycerol, and 20 µM of a unlabeled single strand DNA to mask any binding of Lhr to 32 P-DNA released from the SA-DNA complex. The reaction products were analyzed by electrophoresis through a 15-cm native 15% polyacrylamide gel containing 89 mM Tris-borate, 2.5 mM EDTA. The free 32 P-labeled 34-mer DNA and the SA-DNA complexes were visualized by autoradiography.
Helicase assay. The 5' 32 P-labeled DNA strand was prepared by reaction of a synthetic oligodeoxynucleotide with T4 polynucleotide kinase and [γ 32 P]ATP. The labeled DNA was separated from ATP by gel-filtration through a G-50 micro column (GE Healthcare) and then annealed to a 3-fold excess of a complementary DNA strand to form the various substrates shown in the figures. The annealed DNAs were purified by electrophoresis through a native 12% polyacrylamide gel, eluted from an excised gel slice by incubation for 16 h at 4˚C in 200 µl 10 mM Tris-HCl, pH 7.5, 1 mM EDTA, recovered by ethanol precipitation and resuspended in water. The 5' 32 P-labeled RNA strand was prepared by reaction of a synthetic oligonucleotide with T4 polynucleotide kinase and [γ 32 P]ATP. The labeled RNA was separated from ATP by electrophoresis through a 40-cm 15% polyacrylamide gel containing 7 M urea in 89 mM Tris-borate, 2.5 mM EDTA. The RNA was eluted from an excised gel slice by incubation for 16 h at 4˚C in 1 M ammonium acetate, 0.2% SDS, 20 mM EDTA. The RNA was recovered by ethanol precipitation and resuspended in 10 mM Tris-HCl, pH 6.8, 1 mM EDTA. Annealing of the labeled RNA to complementary strands and gel-purification of the duplexes was performed as described above for the labeled DNA duplexes. Helicase reaction mixtures (10 µl) containing 20 mM Tris-HCl, pH 7.0, 1 mM DTT, 5 mM CaCl 2 , 32 P-labeled nucleic acid as specified, and Lhr as specified were preincubated for 10 min at room temperature. The reactions were initiated by adding 1 mM ATP and a 10-fold excess of an unlabeled oligonucleotide identical to the labeled strand of the helicase substrate. Addition of excess unlabeled strand was necessary to prevent the spontaneous reannealing of the unwound 32 P-labeled strand. The reaction mixtures were incubated for 30 min at 37°C and then quenched by adding 1 µl of a solution containing 5% SDS, 100 mM EDTA. The mixtures were supplemented with 4 µl of 50% glycerol, 0.3% bromophenol blue. The reaction products were analyzed by electrophoresis through a 15-cm 12% polyacrylamide gel in 89 mM Tris-borate, 2.5 mM EDTA. The products were visualized by autoradiography.

Recombinant Lhr is a monomeric DNAdependent ATPase.
To evaluate the enzymatic and physical properties of M. smegmatis Lhr, we produced the full-length 1507-aa protein in E. coli as a His 10 fusion and purified it from a soluble extract by nickel-agarose and DEAE-Sephacel chromatography ( Fig. 2A, WT). We also produced and purified three C-terminal truncation mutants: Lhr-(1-1098), Lhr-(1-856), and Lhr-(1-562), which displayed the expected serial decrements in apparent molecular weight when analyzed by SDS-PAGE ( Fig. 2A). Reaction of wild-type Lhr with 1 mM [α 32 P]ATP in the presence of calcium and single strand DNA resulted in the hydrolysis of [α 32 P]ATP to [α 32 P]ADP, as monitored by PEIcellulose TLC. The extent of ATP hydrolysis was proportional to input Lhr and >90% of the ATP was hydrolyzed at saturating enzyme (Fig. 2B). No conversion of [α 32 P]ADP to [ 32 P]AMP was detected. Whereas the Lhr-(1-1098) and Lhr-(1-856) variants displayed similar ATPase specific activity to that of the wild-type Lhr, the Lhr-(1-562) variant was catalytically inert (Fig. 2B). We conclude that: (i) the C-terminal 651 amino acids are dispensable for ATP phosphohydrolase activity; and (ii) the segment from aa 563 to 856 includes structural elements required for DNAdependent ATP hydrolysis.
The native size of full-length Lhr was analyzed by zonal velocity sedimentation through a 15-30% glycerol gradient. The Lhr preparation was sedimented alone and, in a parallel gradient, as a mixture with marker proteins catalase and BSA. SDS-PAGE analysis of the odd-numbered fractions of the "Lhr-alone" glycerol gradient revealed that the Lhr polypeptide (calculated 165 kDa) sedimented as a single component peaking in fractions 9-11 ( Fig. 3; middle panel). Note that the minor lower molecular weight contaminant polypeptides seen in the DEAE-Sephacel Lhr preparation ( Fig. 2A) were separated from intact Lhr during the sedimentation procedure and were recovered in the "lighter" gradient fractions 13 to 19 (Fig. 3, middle panel). Aliquots of the odd numbered gradient fractions were incubated with 1 mM [α 32 P]ATP in the presence of calcium and single strand DNA and the hydrolysis of [α 32 P]ATP to [α 32 P]ADP was monitored by PEIcellulose TLC. The ATPase activity profile paralleled the abundance of the Lhr protein, peaking in fractions 9 to 11 (Fig. 3, bottom panel). These results indicate that the DNA-dependent ATPase activity is inherent to Lhr.
SDS-PAGE analysis of the odd-numbered fractions of the "Lhr plus markers" glycerol gradient (Fig. 3, top panel) showed that Lhr again sedimented as a discrete peak (fractions 9-11) midway between catalase (peak fraction 5; native size 250 kDa) and BSA (peak fraction 15, native size 66 kDa). We surmise that Lhr is a monomer in solution.
Requirements for the Lhr ATPase activity. The ATPase activity of Lhr in the presence of an 80mer single strand DNA required a divalent cation cofactor (Fig. 4A). When various metals were tested at 1 mM concentration, the divalent cation requirement for ATP hydrolysis was satisfied best by calcium (Fig. 4A). Magnesium and manganese were 40% as effective as calcium; nickel and zinc were 6-9% as effective. Copper and zinc were ineffective (Fig. 4A). Calcium and magnesium titration experiments showed that hydrolysis of 1 mM ATP was optimal at 0.5 to 1 mM (not shown). ATP hydrolysis in the presence of calcium and 80mer DNA was optimal from pH 5.5 to 7.5 in Tris buffer; activity declined sharply at pH 5.0 and was nil at pH 4.5 (Fig. 4B). Activity declined gradually at alkaline pH, to 60% of the peak value at pH 9.5 (Fig. 4B).
NTP substrate specificity. NTP substrate specificity was examined by colorimetric assay of the release of P i from unlabeled ribonucleotides ATP, GTP, CTP or UTP and deoxynucleotides dATP, dGTP, dCTP or dTTP, each at 1 mM concentration. Lhr specifically hydrolyzed ATP and dATP; the other rNTPs and dNTPs were ineffective as substrates (Fig. 4C).
Steady-state kinetics of ATP hydrolysis. We determined steady-state kinetic parameters by measuring the velocity of ATP hydrolysis as a function of ATP concentration in the presence of 1 mM calcium and 5 µM 80-mer single strand DNA cofactor (Fig. 4D). From a nonlinear regression curve fit of the data to the Michaelis-Menten equation, we calculated that Lhr has a K m of 0.10 ± 0.01 mM ATP and a k cat of 6.1 ± 0.2 s -1 .
Lhr translocates unidirectionally 3'-to-5' on single strand DNA. NTP hydrolysis by nucleic acid-dependent phosphohydrolases is often coupled to mechanical work -either duplex unwinding or displacement of protein-nucleic acid complexes -as a consequence of translocation of the phosphohydrolase enzyme along the nucleic acid. To address whether Lhr has translocase activity, we employed a streptavidin displacement assay (12,13,16,(23)(24)(25) as follows. 32 P-labeled 34mer DNA oligonucleotides containing a single biotin moiety at the fourth internucleotide from the 5' end or the second internucleotide from the 3' end were preincubated with excess streptavidin (SA) to form a stable SA-DNA complex that was easily resolved from free biotinylated 34-mer DNA during native PAGE (Fig. 5). The translocation assay scores the motor-dependent displacement of SA from the DNA in the presence of ATP and excess free biotin, which instantly binds to free SA and precludes SA rebinding to the labeled DNA. The rationale of the assay is that directional tracking of the motor along the DNA single strand will displace SA from one DNA end, but not the other. When moving 3' to 5', it can displace SA as it collides with the 5'-biotin-SA. By contrast, a 3'biotin-SA is not expected to be displaced by a 3'to-5' translocase, because the protein moves away from the SA and simply falls off the free 5' end (Fig. 5). The converse outcomes apply to a 5'-to-3' translocase; it displaces a 3' SA, but not a 5' SA. The instructive finding was that Lhr displaced SA from a 5'-biotin-SA complex on the 34-mer single strand DNA to yield the free 32 P-labeled 34-mer strand (Fig. 5, lane 13), but was unable to displace SA from a 3'-biotin-SA complex tested in parallel (Fig. 5, lane 5). Stripping of the 5'-biotin-SA complex by Lhr to liberate free DNA depended on ATP and calcium (Fig. 5, lanes 11 and 12). Another instructive finding was that the ATPaseproficient truncated proteins Lhr-(1-1098) and Lhr-(1-856) were active as 3'-to-5' translocases (Fig. 5,  lanes 14 and 15), whereas the ATPase-defective Lhr-(1-562) was not (Fig. 5, lane 16). Control experiments showed that the capacity of the 5' and 3' biotin-modified DNA strands to activate ATP hydrolysis by Lhr was not impeded by the inclusion of excess SA in the reaction mixtures (Supplemental Fig. S1). We surmise from this result that a 3'-biotin-SA complex does not inhibit the loading of Lhr on the DNA strand, fortifying our inference from the SA displacement experiments that Lhr translocates 3'-to-5' but not 5'-to-3'.
Lhr is a 3'-to-5' DNA helicase. In light of the 3'to-5' translocase activity demonstrated above, we tested Lhr for helicase activity with a 3'-tailed duplex substrate consisting of a 31-bp duplex with a 28-nt 3' single strand tail to serve as a potential "loading strand" (Fig. 6). The helicase assay format we used entailed preincubation of Lhr and labeled DNA, followed by initiation of unwinding by addition of ATP, with simultaneous addition of a "trap" of excess unlabeled 59-mer displaced strand that: (i) minimizes reannealing of any 32 Plabeled 59-mer that was unwound by Lhr and (ii) competes with the loading strand for binding to any free Lhr or Lhr that dissociated from the labeled DNA without unwinding it. Consequently, the assay predominantly gauges a single round of strand displacement by Lhr bound to the labeled 3'tailed duplex prior to the onset of ATP hydrolysis. We found that Lhr unwound the 3'-tailed DNA substrates to yield a radiolabeled free single strand that migrated faster than the input tailed duplex during native PAGE (Fig. 6A); the helicase reaction product comigrated with free 59-mer generated by thermal denaturation of the substrate (Fig. 6A, lane ∆). Lhr failed to unwind a 59-bp blunt duplex DNA substrate (Fig. 6), thereby attesting to a requirement for a single strand tail to serve as a loading strand for the helicase. Lhr also failed to unwind a 5'-tailed duplex substrate (Fig.  6A). Lhr did unwind a "Y fork" duplex consisting of a 31-bp duplex with one blunt end and 28-nt 5' and 3' single strand tails at the other end (Fig. 6A). Assaying the Lhr glycerol gradient fractions for DNA unwinding verified that the helicase activity cosedimented with Lhr in fractions 9-11 (Fig. 6B). These results establish that Lhr is a unidirectional motor, powered by ATP hydrolysis, that tracks 3'to-5' along the loading strand and unwinds duplex DNA.
Single-strand DNA length requirement for ATP hydrolysis. We tested the ability of single strand DNAs of varying length to activate ATP hydrolysis by Lhr. Titration of the oligonucleotides revealed a hyperbolic dependence of ATP hydrolysis on the amount of 59-mer, 24-mer, 18mer, or 12-mer strands (Fig. 7A). Nonlinear regression fitting of the data to a one-site binding model in Prism yielded apparent K d values as follows: 9.4 ± 0.9 nM 59-mer; 11 ± 1.5 nM 24mer; 27 ± 2.6 nM 18-mer; and 210 ± 45 nM 12mer. By contrast, a 6-mer oligonucleotide was ineffective at up to 2 µM concentration (Fig. 7A).
Tail length requirement for duplex DNA unwinding. To probe the 3' tail requirement for duplex unwinding by Lhr, we tested a series of helicase substrates with 3'-oligo(dT) tails of varying length (20, 15, 10, or 5 nucleotides) attached to an identical 24-bp duplex segment (Fig.  7B). Lhr-mediated displacement of the labeled 24mer strand was equivalent when the 3' tail was 15 or 20 nucleotides; helicase specific activity decreased by a factor of six when the 3' tail was retracted to 10 nucleotides (Fig. 7B). Virtually no unwinding was seen when the tail was shortened to only 5 nucleotides. The loading strand 3' tail length requirements paralleled the length requirements for activation of the DNA-dependent ATPase (Fig.  7A).
Activation of ATP hydrolysis by RNA. DNA and RNA strands of otherwise identical size (24-mer) and nucleobase sequence (excepting U for T in RNA) were tested as cofactors for the Lhr ATPase. The dependence of ATP hydrolysis on the level of input DNA or RNA is shown in Fig. 8A. The apparent K d values for the 24-mer DNA and RNA, obtained by nonlinear regression curve fitting of the data in Prism, were 13 ± 1.9 nM and 150 ± 20 nM, respectively. (The calculated activity plateau for the RNA-dependent ATPase reaction was ~85% of the DNA-dependent reaction.) Lhr unwinds an RNA:DNA hybrid with DNA as the loading strand. The RNA-dependent ATPase activity of Lhr raises the question of whether the motor can unwind RNA-containing duplexes. To address this issue, we prepared a series of helicase substrates, consisting of a 24-bp duplex and a 15mer 3' single strand tail, in which the 39-mer loading strand was DNA or RNA of otherwise identical nucleobase sequence (excepting U for T in RNA) and the 24-mer displaced strand was DNA or RNA of identical nucleobase sequence (Fig. 8B). The instructive findings from an initial assay of unwinding of 50 nM duplex by 250 nM Lhr were as follows: (i) an RNA:DNA hybrid with a DNA loading strand was a better helicase substrate than a DNA:DNA duplex; (ii) Lhr failed to unwind a 3'-tailed RNA:RNA duplex; and (iii) Lhr did not unwind a DNA:RNA hybrid with a 39mer RNA loading strand (Fig. 8B). Lhr titration experiments showed that the specific activity of the full-length enzyme in unwinding the RNA:DNA hybrid (Fig. 9B) was 8-fold greater than its activity in unwinding the DNA:DNA duplex (Fig. 9A).

Distinctive effects of C-terminal truncations on DNA:DNA versus RNA:DNA unwinding.
Deletion of the C-terminal 409 or 651 amino acids of Lhr had little impact on the ATPase activity of Lhr (Fig. 2B) and did not impair its 3'-to-5' translocase activity on single strand DNA, as gauged by streptavidin displacement (Fig. 5). By assaying the dependence of ATP hydrolysis on the level of input 24-mer DNA, we found that the apparent K d values of Lhr-(1-1098), and Lhr-(1-856) for the 24-mer DNA were 12 nM and 10 nM, respectively (data not shown), signifying that the C-terminal deletions did not diminish the affinity of Lhr for the DNA cofactor for ATP hydrolysis. However, when we compared the DNA:DNA unwinding activities of Lhr-(1-1098) and Lhr-(1-856) to that of full-length Lhr, we noted that the Cterminal deletions reduced helicase specific activity to 14% and 9% of the wild-type activity, respectively (Fig. 9A). By contrast, the same Cterminal deletions had no impact on the specific activity of Lhr as an RNA:DNA helicase (Fig. 9B). These results signify that the C-terminal 409-aa segment enables of Lhr to translocate through a DNA:DNA duplex, but is dispensable for its translocation through an RNA:DNA hybrid.

DISCUSSION
The present study unveils mycobacterial Lhr as the founding member of a novel clade of SF2 helicases, by virtue of its distinctive biochemical specificities and its signature domain organization. Lhr is a monomeric nucleic acid-dependent ATPase that uses the energy of nucleotide hydrolysis to drive 3'-to-5' translocation along single strand DNA and to unwind duplex DNA or an RNA:DNA hybrid en route. The ATPase is more active in the presence of calcium than magnesium, a feature that distinguishes Lhr from previously characterized mycobacterial helicases.
[To wit, UvrD1 is inactive when magnesium is replaced by calcium (4); RqlH is more active with magnesium than calcium (12); and SftH is equally active with magnesium and calcium (13).] Lhr has no appreciable ATPase activity in the absence of nucleic acid; its phosphohydrolase function is triggered by either single strand DNA or single strand RNA, yet the apparent affinity for the DNA activator is 11-fold higher than for an RNA strand of identical size and nucleobase sequence. The remarkable property of Lhr is that it is 8-fold better at unwinding an RNA:DNA hybrid duplex when translocating on a 3'-tailed DNA loading strand than it is at displacing a DNA:DNA duplex of identical nucleobase sequence. Yet, Lhr is unable to unwind a duplex when the loading strand is RNA.
To score in the single-turnover helicase assay, the motor must initiate ATP hydrolysis and directional translocation on the 3' single strand tail of the loading strand and advance into the duplex segment, prying the strands apart as it moves, to the point that either the non-loading strand is traversed completely and thereby ejected or to a position in the duplex region where the level of residual base-pairing is low enough for the strand to dissociate spontaneously. Thus, the inability of Lhr to unwind from an RNA loading strand, despite the efficacy of RNA in activating the ATPase, could signify that ATP hydrolysis is uncoupled, or loosely coupled, to translocation when the effector nucleic acid is RNA (i.e., Lhr dissociates more readily from an RNA strand after ATP hydrolysis). More interesting is the fact that an RNA strand is preferred over a DNA strand for displacement when Lhr translocates along a DNA loading strand. Possible explanations for this result are: (i) that Lhr senses the helical conformation of the duplex at its leading edge and prefers the Aform helix of an RNA:DNA hybrid to the B-form DNA:DNA duplex; or (ii) that Lhr interacts directly with the displaced non-loading strand and RNA is preferred in this regard.
The ATPase, DNA translocase, and RNA:DNA helicase activities of mycobacterial Lhr are encompassed within an N-terminal 856-aa segment of the 1507-aa polypeptide. This autonomous helicase consists of a classical SF2 (DExH-box) motor domain embraced within the N-terminal ~500-amino acids. However, the Lhr-(1-562) fragment containing the motor domain does not suffice for ATP hydrolysis, signifying that the region from aa 563-856 contributes structural elements necessary for DNA-dependent ATP hydrolysis. The Lhr segment from aa 657 to 847 is classified at NCBI as a "pfam08494" domain. This module, of unknown structure or function, is fused in cis to an upstream helicase-like NTPase domain in many biochemically uncharacterized proteins in the NCBI database. Indeed, the abbreviated Lhr-(1-856) helicase enzyme per se exemplifies a large family of homologous polypeptides, all of similar size (~800-900 aa), present in the proteomes of scores of diverse bacterial and archaeal taxa. The salient property of the Lhr-(1-856) helicase illuminated presently is that it is 100-fold more active in unwinding the tailed RNA:DNA hybrid substrate than the tailed DNA:DNA duplex (Fig.  9). [We speculate below on the possible role of Lhr or Lhr-(1-856) helicases in the repair of gaps flanking RNA-primed DNAs.] In our view, the domain organization of fulllength mycobacterial Lhr is sufficiently distinctive to warrant the designation of Lhr as the exemplar of a new SF2 helicase clade. A search of the NCBI databases with M. smegmatis Lhr recovered scores of bacterial Lhr-like proteins, of similar length (1400 to 1700 aa), and containing a C-terminal segment homologous to the inessential portion of mycobacterial Lhr downstream of aa 856. The Cdomain of mycobacterial Lhr (aa 856-1507), though not important for RNA:DNA unwinding, apparently does enable the DNA:DNA helicase activity. We find no functionally instructive motifs within this portion of mycobacterial Lhr. Homologs of full-length Lhr that include the Cdomain are found in diverse bacterial taxa from 8 different phyla (Table S1). They are especially prevalent in Actinobacteria (in more than 50 genera) and Proteobacteria (more than 25 genera) (Table S1). We did not find "full-length" Lhr homologs in archaea.
Two independent studies of the transcriptional response of M. tuberculosis to DNA damaging agents have identified the Rv3296 gene (encoding the Lhr helicase) as being upregulated in bacteria exposed to UV irradiation or mitomycin C (26,27). The genes encoding helicases AdnAB (Rv3201c Rv3202c) and UvrD2 (Rv3198c) are also upregulated in response to DNA damage (26,27). Such studies implicate Lhr in DNA repair or recombination. Given that Lhr is most adept at translocating 3'-to-5' after loading on singlestranded DNA and then displacing an RNA strand annealed to the tracking strand, we speculate that Lhr might participate in the repair of gaps flanking RNA-primed Okazaki fragments, e.g., by peeling off the 5' RNA segments and exposing them to a flap endonuclease-like processing enzyme. This echoes the suggestion by Steven Matson that the E.      8 and 16). Enzyme was omitted from control reactions in lanes 2 and 10. Enzyme and streptavidin were both omitted from control reactions in lanes 1 and 9. ATP was omitted from the reactions in lanes 3 and 11. Calcium was omitted and 5 mM EDTA was added to the reactions in lanes 4 and 12.