Abstract
The evolutionary conserved PIF1 DNA helicase family appears to have largely nonoverlapping cellular functions. To better understand the functions of human PIF1, we investigated biochemical properties of this protein. Analysis of single-stranded (ss) DNA-dependent ATPase activity revealed nonstructural ssDNA to greatly stimulate ATPase activity due to a high affinity for PIF1, even though PIF1 preferentially unwinds forked substrates. This suggests that PIF1 needs a ssDNA region for loading and a forked structure for translocation entrance into a double strand region. Deletion analysis demonstrated novel functions of a unique N-terminal portion, named the PIF1 N-terminal (PINT) domain. When the PINT domain was truncated, apparent affinity for ssDNA and unwinding activity were much reduced, even though the maximum velocity of ATPase activity and Km value for ATP were not affected. We suggest that the PINT domain contributes to enhancing the interaction with ssDNA through intrinsic binding activity. In addition, we found DNA strand-annealing activity, also residing in the PINT domain. Notably, the unwinding and annealing activities were inhibited by replication protein A. These results suggest that the functions of PIF1 might be restricted with particular situations and DNA structures.
INTRODUCTION
Helicases are ubiquitous enzymes that catalyze the unwinding of DNA duplexes using ATP as their energy source. They therefore play vital roles in nearly all DNA metabolic processes, including DNA replication, recombination and repair. The PIF1 subfamily of 5′–3′ DNA helicases ( 1–10 ) belongs to the SFI superfamily, conserved in diverse organisms ( 11 ).
In Saccharomyces cerevisiae ( Sc ), Sc Pif1 was originally identified because of its involvement in recombination of mitochondrial DNA (mtDNA) ( 12 , 13 ). Dysfunction of Sc Pif1 leads to mitochondrial genetic instability due to spontaneous oxidative damage ( 14–16 ), and induces mtDNA damage ( 17 ). Sc Pif1p has not only mitochondrial targeting but also nuclear targeting signals and therefore is localized in both the mitochondria and nucleus ( 18 ). Nuclear Sc Pif1p has multiple functions. When it is over-produced, telomeres become shorter, while they elongate when it is eliminated ( 18 , 19 ). In the absence of nuclear Sc Pif1, gross chromosome rearrangement is also increased, and healing of double-stranded broken ends via telomere addition increases ∼200- to 1000-fold. These data suggest a negative regulatory role of Sc Pif1 in telomere metabolism ( 18–24 ). Indeed, Sc Pif1 catalytically inhibits telomerase activity in vitro ( 25 ). Other genetic data suggest that Sc Pif1 plays roles in Okazaki fragment processing ( 26 ), pausing of replication progression at ribosomal DNA loci ( 27 ) and unwinding of hemicatenans ( 28 ).
Schizosaccharomyces pombe encodes a PIF1 homolog, Pfh1 ( 8 , 10 ) which is required for cell cycle progression in late S -phase and for appropriate responses to DNA damage agents ( 8 , 10 ). It is also implicated in lagging strand DNA processing ( 7 ).
Previous biochemical studies of human PIF1 were performed using N-terminal-truncated forms of PIF1, containing only the conserved helicase motifs, located in the C-terminus, since it earlier proved impossible to obtain purified full-length human PIF1 protein ( 3 , 4 , 9 ). However, we found that the N-terminal region of human PIF1, named here as the PIF1 N-terminal (PINT) domain, is well conserved in the PIF1 family, suggesting a possible functional role. Here, we established a method to purify full-length human PIF1 protein and provided the first evidence that the PINT domain has crucial functions in this enzyme.
MATERIALS AND METHODS
Plasmid construction
The nucleotide sequence of the 5′-end of human PIF1 cDNA was obtained by 5′-rapid amplification of cDNA ends. Then the full-length human PIF1 cDNA was amplified from a HeLa cDNA library using primers, CATATGCTCTCGGGCATAGAGGCGGCGGCAGGGGAATATGAGGACTCG and TCAGAGGATTGGGTCCATGTT by PCR, then the nucleotide sequences of the clones were verified and submitted to the database with the accession number, EU084033. The entire coding region with a histidine tag at the N-terminus was inserted into the pET20b(+) vector to yield pET20b-PIF1. The truncated forms, PIF 167–641 and PIF 1–180 consisting of the numbered amino acid residues were also cloned into pET20b(+) and pET15b, respectively, to produce his-tagged fusion proteins. The structures of the resultant plasmids, pET20b-PIF1, pET20b-PIF 167–641 and pET15b-PIF 1–180 , are shown in Supplementary Figure S1 . In this article, PIF 167–641 and PIF 1–180 are referred to as C-terminal region of PIF1 (PIF1C) and N-terminal region of PIF1 (PIF1N), respectively.
Protein purification
RPA was purified as described ( 29 ) from over producing Escherichia coli cells ( 30 ). PIF1 and its deletion derivatives were purified as his-tagged fusion proteins at the N-termini. During all the purification steps, induced proteins were monitored by SDS–PAGE followed by staining with Coomassie Brilliant Blue R-250, or western blotting using Penta-His antibody (#34660, QIAGEN, Tokyo, Japan) or anti-PIF1 antibodies. Protein concentrations were determined by Bio-Rad protein assay using BSA (Bio-Rad, Tokyo, Japan) as the standard.
His-tagged full-length PIF1 and PIF1C were purified from overexpressing E. coli cells, BL21 (DE3) ( 31 ). The strain harboring a plasmid pMStRNA1, in which tRNAs for rare codons were cloned into a R6K derived kanamycin resistant plasmid ( 32 ), and pET20b-PIF1 was grown in 3 l of LB supplemented with ampicillin (250 μg/ml) and kanamycine (30 μg/ml) at 15°C, with aeration until the culture reached an A 600 value of 0.6. Isopropyl β- d -thiogalactopyranoside (IPTG) was added to 0.2 mM, and the incubation was continued for 14 h. The resultant cell paste (9 g) was resuspended in 18 ml of buffer I (50 mM HEPES NaOH pH 7.5, 0.1 mM EDTA, 10 mM β-mercaptoethanol, 1 M NaCl) and frozen in liquid nitrogen. The cells were thawed in ice water and lyzed by addition of 3 ml buffer I containing 100 mM spermidine and 4 mg/ml lysozyme. After incubation on ice for 30 min, heating in a 37°C water bath for 2 min and further incubation on ice for 30 min, the lyzate was clarified by centrifugation twice at 85 000 g for 30 min at 4°C. Subsequent column chromatography was carried out at 4°C using a fast protein liquid chromatography (FPLC) system (GE Healthcare, Tokyo, Japan). After adding imidazole to 50 mM, the lyzate was applied at 0.2 ml/min to a 1-ml HiTrap chelating column (GE Healthcare), which had been treated with 0.1 M NiSO 4 and then equilibrated with buffer A (50 mM HEPES NaOH pH 7.5, 10% glycerol, 10 mM β-mercaptoethanol, 1 M NaCl) containing 50 mM imidazole. The column was washed with 10 ml of equilibration buffer at 0.2 ml/min and his-tagged PIF1 was eluted with 10 ml of buffer A containing 100 mM imidazole. Fractions eluted with 100 mM imidazole were pooled and diluted to 50 mM imidazole with buffer A, then loaded again onto a 1-ml HiTrap chelating column at 0.2 ml/min. The column was washed, and PIF1 was eluted with buffer A containing 300 mM imidazole, then loaded at 0.1 ml/min onto a Superdex 200 10/300 GL column (GE Healthcare) equilibrated with buffer A. PIF1 peak fractions were pooled, frozen in liquid nitrogen, and stored at –80°C. His-tagged PIF1C was purified under the same conditions as described for his-tagged PIF1.
His-tagged human PIF1N was purified from overexpressing E. coli cells, Rosetta 2 (DE3) (Novagen, Tokyo, Japan). The strain harboring pET15-PIF1N was grown in 3 l of LB supplemented with ampicillin (250 μg/ml) and chloramphenicol (30 μg/ml) at 15°C with aeration until the culture reached an A 600 value of 0.6. IPTG was added to 0.2 mM, the incubation was continued for 14 h, and the cells were lyzated as described. After adding imidazole to 50 mM, the lyzate was applied at 0.2 ml/min to a 1-ml HiTrap chelating column, which had been treated with 0.1 M NiSO 4 and then equilibrated with buffer A containing 50 mM imidazole. The column was washed with 10 ml of equilibration buffer at 0.2 ml/min and then with 10 ml of buffer A containing 100 mM imidazole. His-tagged PIF1N was eluted with 10 ml of 300 mM imidazole in buffer A. Fractions containing PIF1N were pooled, diluted with buffer B (50 mM HEPES NaOH pH7.5, 10 mM β-mercaptoethanol) to 100 mM of NaCl, and applied at 0.5 ml/min to a 1-ml HiTrap SP HP column (GE Healthcare) equilibrated with buffer B containing 100 mM NaCl. The column was washed with 10 ml of equilibration buffer at 0.1 ml/min, and the PIF1N was eluted with 20 ml of a linear gradient of 100–1000 mM NaCl in buffer B. Fractions containing PIF1N were pooled, frozen in liquid nitrogen and stored at –80°C.
Antibodies
To obtain polyclonal antibodies against PIF1, truncated his-tagged PIF1 proteins (1–180 and 338–641 amino acids) were expressed in Rosetta 2 (DE3), purified and used to immunize rabbits.
DNA substrates
The oligonucleotides employed for the preparation of DNA substrates are listed in Table 1 . Oligonucleotides were 5′-end labeled using [γ- 32 P]ATP (GE Healthcare) and polynucleotide kinase (New England BioLabs, Tokyo, Japan). The schematic structures of substrates are shown in figures, and the labeled oligonucleotides are indicated with asterisks. Annealing reaction mixtures (30 μl) containing the 5′- 32 P-labeled oligonucleotides at 1 μM, all unlabeled oligonucleotides at 3 μM, 10 mM Tris–HCl (pH 7.5), 7 mM MgCl 2 and 200 mM NaCl were heated at 95°C for 10 min, transferred directly to 65°C and held at that temperature for 1 h, slow-cooled to 25°C over a period of 2 h and held at that temperature for 30 min and then cooled to 4°C. Substrates were then purified by electrophoresis through 15–25% polyacrylamide using 0.5× TBE ( 33 ) as the electrophoresis buffer. Substrates were eluted from the gel by crushing the gel slice in TE buffer and incubating overnight at 4°C. The slurry was then filtered through Micro Bio-Spin (Bio-Rad) columns, and the DNA was recovered by ethanol precipitation and resuspended in TE buffer.
ATPase assays
ATPase activity was measured in a standard reaction mixture (20 μl) containing 50 mM Tris–HCl (pH 8.0), 2 mM DTT, 1.2 mM MgCl 2 , 0.25 mg/ml BSA, 2 mM [γ- 32 P]ATP, indicated DNA and 1 μl of protein sample diluted with buffer D (50 mM Tris-HCI pH8.0, 1 M NaCl, 2 mM DTT, 10% glycerol, 0.1 mg/ml BSA) to obtain indicated final concentrations. After preincubation for 30 s at 30°C, reactions were initiated by the addition of PIF1 proteins and further incubated for 10 min. After the reaction was stopped with 4 μl of 20 mM EDTA (pH 8.0), an aliquot (2 μl) was spotted onto a polyethyleneimine-cellulose plate (Merck, Tokyo, Japan) and developed in 0.3 M LiCl/0.9 M formic acid. The products were analyzed using a Bio-Imaging Analyzer BAS2000 (Fuji Photo Film Co., Ltd., Tokyo, Japan). The extents of ATP hydrolysis were measured with reference to the relative ratios of radioactivity of inorganic phosphate to uncleaved ATP.
Kinetic assays to determine Km values for ATP were performed for 10 min in 20 μl reaction mixtures using 14 nM of PIF1 and PIF1C with 3.8 μM and 150 μM (in nucleotides) of M13 mp7 single-stranded (ss) DNA, respectively. Concentrations of ATP ranged from 25 to 400 μM. Km values were evaluated from the plot of the initial velocity versus the ATP concentration using a hyperbolic curve-fitting program with correlation coefficients ( R2 ) >0.99.
DNA helicase assays
DNA helicase activity was measured under ATPase assay conditions (20 μl) with the indicated DNA substrate (0.35 nM) and 1 μl of protein sample diluted with buffer D to obtain the indicated final concentrations. After preincubation for 30 s at 30°C, reactions were initiated by the addition of PIF1 proteins and incubated for 10 min at 30°C. Helicase reactions were terminated with 10 μl of stop solution (150 mM EDTA, 30% glycerol, 2% SDS, 0.1% bromophenol blue). In kinetic experiments, a 160-μl reaction mixture was incubated at 30°C and 10-μl aliquots were withdrawn at the indicated times. The reaction products were subjected to electrophoresis through 15–25% polyacrylamide gels in Tris–glycine buffer ( 33 ). The gels were dried on DE81 paper (Whatman, Tokyo, Japan) and autoradiographed. DNA products were quantified using a Bio-Imaging Analyzer.
Electrophoretic mobility shift assays
Oligonucleotides were labeled with polynucleotide kinase (New England BioLabs) and [γ- 32 P]ATP. Assays of DNA binding were performed with a modification of a method described previously ( 34 ). The reactions (20 μl) were carried out under ATPase assay conditions, sometimes omitting MgCl 2 or ATP, with 25 pM oligonucleotides and 1 μl of protein sample diluted with buffer D to obtain the indicated final concentrations. Incubation was carried out on ice for 10 min followed by loading on prerunning 5% polyacrylamide gels (79:1 acrylamide/ bis -acrylamide). The electrophoresis buffer contained 6 mM Tris–HCl (pH 8.0), 5 mM sodium acetate and 1 mM EDTA, and the gels were subjected to a constant voltage of 8 V/cm for 100 min at 4°C. Following gel electrophoresis, the products were analyzed as described for the helicase assay. For quantification, fractions of free DNA were measured, and the binding fractions were determined by subtraction from the amount of the free DNA at 0 nM of the protein ( 35 ).
For RPA binding experiments, the reactions (20 μl) were carried out under the ATPase assay conditions, sometimes omitting ATP. The substrate 3F:4L (0.35 nM) was incubated, directly or after heating at 100°C for 5 min, with 1 μl of RPA sample diluted with buffer (50 mM HEPES NaOH pH 7.5, 250 mM NaCl, 10 mM β-mercaptoethanol, 10% glycerol) to obtain indicated final concentrations. Incubation was carried out on ice for 10 min, and the products were analyzed as in the PIF1 binding experiments.
DNA strand annealing assays
Strand annealing reactions (20 μl) were carried out under the ATPase assay conditions, but in the presence of the indicated concentrations of ATP, with the 5′-end labeled substrate DNA for helicase assay (0.35 nM), which had been boiled at 100°C for 5 min and quickly chilled on ice before adding to the reaction mixture, and 1 μl of protein sample diluted with buffer D to obtain the indicated final concentrations. After preincubation for 30 s at 30°C, reactions were initiated by the addition of PIF1 proteins, with incubation for 10 min at 30°C. After terminating the reactions with 10 μl of stop solution, the products were analyzed as described for the helicase assay. Kinetic experiments were also carried out as described for the helicase assay.
RESULTS
Purification of PIF1 protein and its deletion derivatives
Previously, biochemical studies of human PIF1 were performed using N-terminal-truncated forms of PIF1 containing only the conserved helicase motifs, located in the C-terminus ( Supplementary Figure S2A ). However, we found that the PINT domain is well conserved in the PIF1 family ( Supplementary Figure S2B ), suggesting that it plays a functional role. To examine biochemical activity, we established a procedure to purify full-length PIF1 with a 6× histidine tag at the N-terminus at quantities sufficient for detailed biochemical studies from overproducing E. coli cells ( Supplementary Figure S2C ). We also purified a C-terminal truncated form (PIF1N) and a N-terminal truncated form (PIF1C) ( Supplementary Figure S2A ) to address biochemical functions of individual domains. PIF1C and PIF1N consist of only the seven helicase motifs and only the PINT domain, respectively ( Supplementary Figure S2A and Materials and methods section). Elution profiles from gel filtration chromatography suggested PIF1 and PIF1C to be monomers (data not shown) as described in yeast homologs ( 1 , 5 , 7 ). The purified proteins were analyzed by SDS–PAGE followed by CBB staining and western blotting, showing PIF1, PIF1C and PIF1N to have molecular sizes of 71, 54 and 22 kDa, respectively ( Supplementary Figure S2C ).
Unwinding activities of PIF1 and PIF1C, a mutant lacking PINT domain but containing the helicase domain
We first measured DNA helicase activity of the purified proteins using the indicated DNA substrate ( Figure 1 A). When 5′ overhang (1F:2S) and 3′ overhang (1R:2S) DNA were used as substrates, we detected helicase activity only on the 5′-overhang substrate, consistent with previous reports ( 3 , 9 ), although the activity was very low ( Figure 1 B). In the titration experiment, activity was only detected clearly at the maximum concentration of the purified sample ( Figure 1 B). Since yeast PIF1 homologs preferentially unwind forked structures ( 1 , 5 , 7 ), we tested a forked substrate ( Figure 1 C). The titration experiment demonstrated that human PIF1 efficiently unwound the forked substrate with about 10 times higher activity than that for the 5′-overhang substrate ( Figure 1 C), suggesting that the property was conserved in evolution. Time course experiments using the forked substrate demonstrated that PIF1 could unwind up to 50% of the substrate in a 15 min reaction, although longer incubation did not increase the products ( Figure 1 D).
Helicase activity of PIF1. (A) Schematic structures of the DNA substrates. The asterisks indicate 32 P-labeled 5′ phosphate. (B) Helicase activity of PIF1. Increasing levels of PIF1 were incubated with substrates (0.35 nM) with the 5′overhang, 1F:2S (upper panel), or 3′ overhang, 1R:2S (lower panel) under standard reaction mixtures at 30°C for 10 min. Reaction products were separated on a 15–25% polyacrylamide gel. (C) Helicase activity of PIF1, PIF1N and PIF1C proteins. The forked structure partial duplex DNA substrate, 3F:4L (0.35 nM), was incubated with the indicated concentrations of PIF1, PIF1N and PIF1C under standard reaction conditions at 30°C for 10 min. The quantified data are shown graphically. The errors in the experiments were <10%. (D) Time course of unwinding reactions with PIF1. The forked structure partial duplex DNA substrate, 3F:4L (0.35 nM), was incubated at 30°C for the indicated time under standard reaction conditions with PIF1 (33 nM). The quantified data are shown graphically. The errors in the experiments were <10%.
Helicase activity of PIF1. (A) Schematic structures of the DNA substrates. The asterisks indicate 32 P-labeled 5′ phosphate. (B) Helicase activity of PIF1. Increasing levels of PIF1 were incubated with substrates (0.35 nM) with the 5′overhang, 1F:2S (upper panel), or 3′ overhang, 1R:2S (lower panel) under standard reaction mixtures at 30°C for 10 min. Reaction products were separated on a 15–25% polyacrylamide gel. (C) Helicase activity of PIF1, PIF1N and PIF1C proteins. The forked structure partial duplex DNA substrate, 3F:4L (0.35 nM), was incubated with the indicated concentrations of PIF1, PIF1N and PIF1C under standard reaction conditions at 30°C for 10 min. The quantified data are shown graphically. The errors in the experiments were <10%. (D) Time course of unwinding reactions with PIF1. The forked structure partial duplex DNA substrate, 3F:4L (0.35 nM), was incubated at 30°C for the indicated time under standard reaction conditions with PIF1 (33 nM). The quantified data are shown graphically. The errors in the experiments were <10%.
To compare the unwinding activity of PIF1C, titration experiments were performed with the forked substrate. We found that PIF1C exhibited helicase activity, but it required more than 100 times more protein to obtain equivalent activity to that of PIF1 ( Figure 1 C). As a control, we showed that PINT domain itself could not unwind the substrate ( Figure 1 C). These results suggested that the function of the PINT domain could be enhancement of the unwinding activity of the helicase domain.
Nonstructural ssDNA, but not forked-structural DNA, preferentially stimulates ATPase activity of PIF1
DNA helicases are enzymes with an associated DNA-dependent ATPase activity. They are presumed to use the hydrolysis of ATP to translocate to ssDNA and to subsequently break the hydrogen bonds of duplex DNA. Characterization of ATPase activity could provide important information as a helicase. To analyze ssDNA dependent ATPase activity of PIF1, titration of M13 mp7 ssDNA was performed in reactions with optimal concentrations of ATP and MgCl 2 ( Figure 2 A). The result showed that the ATPase activity was increased depending on the concentrations of ssDNA and reached a plateau between 3 and 50 μM (in nucleotide equivalents) of ssDNA ( Figure 2 A). The maximum rate of ATP hydrolysis was calculated to be about 1 000 min −1 , which was equivalent to 5000 min −1 at 37°C of Sc Pif1 ( 5 ) and 4000 min −1 at 30°C of the fission yeast homolog, Pfh1 ( 7 ), which have been determined under reaction conditions with nearly saturated concentrations of ssDNA.
Effects of various DNA molecules on stimulation of ATPase activity of PIF1. (A) Titration of M13 mp7 ssDNA for its stimulation of ATPase activity of PIF1, PIF1N and PIF1C. ATPase activities were measured under standard reaction conditions with PIF1 proteins (13.9 nM) at 30°C for 10 min. An aliquot (2 μl) of the reaction products was then analyzed by thin layer chromatography. The extent of ATP hydrolysis was measured by determining the relative radioactivity of the released inorganic phosphate. Inset is a plot of the data at lower concentrations of ss DNA. The errors in the experiments were <10%. (B) Effects of the various DNA molecules on stimulation of ATPase activity of PIF1. Assays were performed with the indicated DNA (3.8 μM in nucleotides) and PIF1 (13.9 nM). (C) Effects of oligonucleotides with different compositions on stimulation of ATPase activity by PIF1. Assays were performed with indicated DNA (7.5 μM in nucleotides) and PIF1 (14.6 nM). (D) Effects of oligonucleotides with different compositions on stimulation of ATPase activity by PIF1C. Assays were performed with indicated DNA (150 μM in nucleotides), and PIF1C (13.9 nM).
Effects of various DNA molecules on stimulation of ATPase activity of PIF1. (A) Titration of M13 mp7 ssDNA for its stimulation of ATPase activity of PIF1, PIF1N and PIF1C. ATPase activities were measured under standard reaction conditions with PIF1 proteins (13.9 nM) at 30°C for 10 min. An aliquot (2 μl) of the reaction products was then analyzed by thin layer chromatography. The extent of ATP hydrolysis was measured by determining the relative radioactivity of the released inorganic phosphate. Inset is a plot of the data at lower concentrations of ss DNA. The errors in the experiments were <10%. (B) Effects of the various DNA molecules on stimulation of ATPase activity of PIF1. Assays were performed with the indicated DNA (3.8 μM in nucleotides) and PIF1 (13.9 nM). (C) Effects of oligonucleotides with different compositions on stimulation of ATPase activity by PIF1. Assays were performed with indicated DNA (7.5 μM in nucleotides) and PIF1 (14.6 nM). (D) Effects of oligonucleotides with different compositions on stimulation of ATPase activity by PIF1C. Assays were performed with indicated DNA (150 μM in nucleotides), and PIF1C (13.9 nM).
To describe precisely the concentration of ssDNA required for ATPase activity, a kinetic parameter, Keff , defined as the concentration of ssDNA required to achieve half-maximal ATP hydrolysis ( 5 , 36 , 37 ) was calculated from the titration curve ( Figure 2 A) using a hyperbolic curve-fitting program ( Table 2 ). The calculated Keff value, 0.35 μM (in nucleotides) for M13 mp18 ssDNA, agreed with the 0.6 μM (in nucleotides) of Sc Pif1 reported for the same M13 ssDNA ( 5 ).
Kinetic parameters for the ATPase activity of PIF1 with various ssDNAs
| DNA | PIF1 Keffa (μM) | PIF1C Keff (μM) |
|---|---|---|
| M13 mp7 ssDNA | 0.35 | 8.7 |
| M13 mp18 ssDNA | 0.45 | – b |
| M13 mp18 dsDNA | 9.7 | – |
| 1F c | 0.17 | – |
| hairpin ( 18 ) c | 0.71 | – |
| fork (5:7) c | 0.39 | – |
| p/t (90:30) c | 0.80 | – |
| d(AC)60 c | 0.13 | 3.7 |
| d(CT)60 c | 0.11 | 2.3 |
| d(TG)30 c | 0.12 | 3.4 |
| dT60 c | 0.11 | 6.4 |
| dC60 c | 0.07 | 2.3 |
| dA60 c | 1.2 | nd d |
| d(AG)60 c | 1.8 | nd |
| DNA | PIF1 Keffa (μM) | PIF1C Keff (μM) |
|---|---|---|
| M13 mp7 ssDNA | 0.35 | 8.7 |
| M13 mp18 ssDNA | 0.45 | – b |
| M13 mp18 dsDNA | 9.7 | – |
| 1F c | 0.17 | – |
| hairpin ( 18 ) c | 0.71 | – |
| fork (5:7) c | 0.39 | – |
| p/t (90:30) c | 0.80 | – |
| d(AC)60 c | 0.13 | 3.7 |
| d(CT)60 c | 0.11 | 2.3 |
| d(TG)30 c | 0.12 | 3.4 |
| dT60 c | 0.11 | 6.4 |
| dC60 c | 0.07 | 2.3 |
| dA60 c | 1.2 | nd d |
| d(AG)60 c | 1.8 | nd |
Kinetic assays to determine Keff values were performed for 10 min in 20 μl reaction mixtures using 14 nM of PIF1 and PIF1C. Concentrations of ssDNA ranged from 0.15 to 300 μM in nucleotides. Keff values were evaluated from the plot of the initial velocity versus the nucleotide concentration using a hyperbolic curve-fitting program with the correlation coefficients ( R2 ) more than 0.97, which were shown in supplementary Figure S3 .
aKeff is expressed in nucleotides equivalents.
b Not determined.
c Nucleotide sequences were shown in Table 1 .
d Not detected.
Since PIF1 preferentially unwound the forked substrate ( Figure 1 ), we consider that substrate specificity for unwinding activity could be correlated with stimulation of ATPase activity. To seek preferential structures for stimulation for ATPase of PIF1, we tested various DNAs, including a fork-structure, a primer-template-structure, a hairpin-structure and a linear oligonucleotide ( Table 1 ), as well as M13 mp18 ssDNA and double stranded (ds) DNA ( Figure 2 B). The reactions were carried out with 3.8 μM (in nucleotides) of each oligonucleotide, which was nearly saturated concentration for M13 mp7 ssDNA ( Figure 2 A inset). Among these, M13 mp18 ssDNA proved to strongly and dsDNA to poorly stimulate ATPase activity ( Figure 2 B). Moreover, we found surprisingly that a oligonucleotide, 1F, which would not be expected to form secondary structures ( Table 1 ), stimulated ATPase activity to the highest level, equivalent to that of M13 mp18 ssDNA ( Figure 2 B). The result suggested that an ssDNA region itself, rather than the structure of the DNA, is crucial for stimulation of ATPase activity of PIF1. Consequently, we compared ATPase activity stimulated by seven different 60-mer oligonucleotides composed of one or two nucleotides, guaranteed not to form secondary structures. The reactions were performed in the presence of a nearly saturated concentration of the oligonucleotides (7.5 μM in nucleotides) ( Supplementary Figure S3A ). The results revealed general features of ssDNA for stimulation of ATPase ( Figure 2 C), with the order of stimulation being poly(purine–pyrimidine) ≥ polypyrimidine ≫ polypurine.
Synthetic oligonucleotides used in this work
| Name | Sequence |
|---|---|
| 1F (73-mer) | 5 ′ -CTCCTCCTCCTCCTCCTCCTCCTCCTCCTCCTCCTCCTCCTCCTCCTCGCTCGACGCCATTAATAAT GTTTTC-3′ |
| 1R (73-mer) | 5 ′ -GCTCGACGCCATTAATAATGTTTTCCTCCTCCTCCTCCTCCTCCTCCTCCTCCTCCTCCTCCTCCTCCT CCTC-3′ |
| 2L (50-mer) | 5 ′ -GAAAACATTATTAATGGCGTCGAGCTAGGCACAAGGCGAACTGCTAACGG-3 ′ |
| 2S (25-mer) | 5 ′ -GAAAACATTATTAATGGCGTCGAGC-3 ′ |
| 3F (55-mer) | 5 ′ -AGGTCTCGACTAACTCTAGTCGTTGTTCCACCCGTCCACCCGACGCCACCTCCTG-3 ′ |
| 3S (25-mer) | 5 ′ -CCCGTCCACCCGACGCCACCTCCTG-3 ′ |
| 4S (25-mer) | 5 ′ -CAGGAGGTGGCGTCGGGTGGACGGG-3 ′ |
| 4L (51-mer) | 5 ′ -GCAGGAGGTGGCGTCGGGTGGACGGGATTGAAATTTAGGCTGGCACGGTCG-3 ′ |
| 5 (fork) (55-mer) | 5 ′ -TTTTTTTTTTTTTTTTTTTTTTTTTCGGACGCTCGACGCCATTAATAATGTTTTC-3 ′ |
| 7 (fork) (56-mer) | 5 ′ -TGAAAACATTATTAATGGCGTCGAGCGTCCGTTTTTTTTTTTTTTTTTTTTTTTTT-3 ′ |
| 18 (hairpin) (73-mer) | 5 ′ -GCCTGCACGTGGCGATCGTTCGATCGCCACGTGCAGGCGCTGCTGCTGGCTCGACGCCATTAAT AATGTTTTC-3 ′ |
| 90 (p/t) (90-mer) | 5 ′ -TGGGCTCACGTGGTCGACGCTGGAGGTGATCACCAGATGATTGCTAGGCATGCTTTCCGCAA GAGAACGGGCGTCTGCGTACCCGTGCAG-3 ′ |
| 30 (p/t) (30-mer) | 5 ′ -CTGCACGGGTACGCAGACGCCCGTTCTCTT-3 ′ |
| d(AC)60 (60-mer) | 5 ′ -ACACACACACACACACACACACACACACACACACACACACACACACACACACACACACAC-3 ′ |
| d(AG)60 (60-mer) | 5 ′ -AGAGAGAGAGAGAGAGAGAGAGAGAGAGAGAGAGAGAGAGAGAGAGAGAGAGAGAGAGAG-3 ′ |
| d(TG)60 (60-mer) | 5 ′ -TGTGTGTGTGTGTGTGTGTGTGTGTGTGTGTGTGTGTGTGTGTGTGTGTGTGTGTGTGTG-3 ′ |
| d(CT)60 (60-mer) | 5 ′ -CTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCT-3 ′ |
| d(CT)55 (55-mer) | 5 ′ -TCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCT-3 ′ |
| d(CT)45 (45-mer) | 5 ′ -TCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCT-3 ′ |
| d(CT)35 (35-mer) | 5 ′ -TCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCT-3 ′ |
| d(CT)30 (30-mer) | 5 ′ -CTCTCTCTCTCTCTCTCTCTCTCTCTCTCT-3 ′ |
| d(CT)20 (20-mer) | 5 ′ -CTCTCTCTCTCTCTCTCTCT-3 ′ |
| d(CT)15 (15-mer) | 5 ′ -TCTCTCTCTCTCTCT-3 ′ |
| dC60 (60-mer) | 5 ′ -CCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCC-3 ′ |
| dA60 (60-mer) | 5 ′ -AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA-3 ′ |
| dT60 (60-mer) | 5 ′ -TTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTT-3 ′ |
| Name | Sequence |
|---|---|
| 1F (73-mer) | 5 ′ -CTCCTCCTCCTCCTCCTCCTCCTCCTCCTCCTCCTCCTCCTCCTCCTCGCTCGACGCCATTAATAAT GTTTTC-3′ |
| 1R (73-mer) | 5 ′ -GCTCGACGCCATTAATAATGTTTTCCTCCTCCTCCTCCTCCTCCTCCTCCTCCTCCTCCTCCTCCTCCT CCTC-3′ |
| 2L (50-mer) | 5 ′ -GAAAACATTATTAATGGCGTCGAGCTAGGCACAAGGCGAACTGCTAACGG-3 ′ |
| 2S (25-mer) | 5 ′ -GAAAACATTATTAATGGCGTCGAGC-3 ′ |
| 3F (55-mer) | 5 ′ -AGGTCTCGACTAACTCTAGTCGTTGTTCCACCCGTCCACCCGACGCCACCTCCTG-3 ′ |
| 3S (25-mer) | 5 ′ -CCCGTCCACCCGACGCCACCTCCTG-3 ′ |
| 4S (25-mer) | 5 ′ -CAGGAGGTGGCGTCGGGTGGACGGG-3 ′ |
| 4L (51-mer) | 5 ′ -GCAGGAGGTGGCGTCGGGTGGACGGGATTGAAATTTAGGCTGGCACGGTCG-3 ′ |
| 5 (fork) (55-mer) | 5 ′ -TTTTTTTTTTTTTTTTTTTTTTTTTCGGACGCTCGACGCCATTAATAATGTTTTC-3 ′ |
| 7 (fork) (56-mer) | 5 ′ -TGAAAACATTATTAATGGCGTCGAGCGTCCGTTTTTTTTTTTTTTTTTTTTTTTTT-3 ′ |
| 18 (hairpin) (73-mer) | 5 ′ -GCCTGCACGTGGCGATCGTTCGATCGCCACGTGCAGGCGCTGCTGCTGGCTCGACGCCATTAAT AATGTTTTC-3 ′ |
| 90 (p/t) (90-mer) | 5 ′ -TGGGCTCACGTGGTCGACGCTGGAGGTGATCACCAGATGATTGCTAGGCATGCTTTCCGCAA GAGAACGGGCGTCTGCGTACCCGTGCAG-3 ′ |
| 30 (p/t) (30-mer) | 5 ′ -CTGCACGGGTACGCAGACGCCCGTTCTCTT-3 ′ |
| d(AC)60 (60-mer) | 5 ′ -ACACACACACACACACACACACACACACACACACACACACACACACACACACACACACAC-3 ′ |
| d(AG)60 (60-mer) | 5 ′ -AGAGAGAGAGAGAGAGAGAGAGAGAGAGAGAGAGAGAGAGAGAGAGAGAGAGAGAGAGAG-3 ′ |
| d(TG)60 (60-mer) | 5 ′ -TGTGTGTGTGTGTGTGTGTGTGTGTGTGTGTGTGTGTGTGTGTGTGTGTGTGTGTGTGTG-3 ′ |
| d(CT)60 (60-mer) | 5 ′ -CTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCT-3 ′ |
| d(CT)55 (55-mer) | 5 ′ -TCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCT-3 ′ |
| d(CT)45 (45-mer) | 5 ′ -TCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCT-3 ′ |
| d(CT)35 (35-mer) | 5 ′ -TCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCT-3 ′ |
| d(CT)30 (30-mer) | 5 ′ -CTCTCTCTCTCTCTCTCTCTCTCTCTCTCT-3 ′ |
| d(CT)20 (20-mer) | 5 ′ -CTCTCTCTCTCTCTCTCTCT-3 ′ |
| d(CT)15 (15-mer) | 5 ′ -TCTCTCTCTCTCTCT-3 ′ |
| dC60 (60-mer) | 5 ′ -CCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCC-3 ′ |
| dA60 (60-mer) | 5 ′ -AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA-3 ′ |
| dT60 (60-mer) | 5 ′ -TTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTT-3 ′ |
Since these experiments were performed with nearly saturated concentrations of oligonucleotides, the results do not reflect affinities of the respective ssDNAs. We therefore determined a kinetic parameter, Keff , calculated from data of ssDNA-titration experiments ( Supplementary Figure S3A , data not shown) using a hyperbolic curve-fitting program ( Table 2 ). The Keff value for dsDNA was more than 20 times higher than those for ssDNA, showing a preference to ssDNA. For ssDNA, the Keff values, except with polypurine, were essentially the same (∼0.1 μM in nucleotides), and slightly lower than with M13 ssDNA and structured DNAs, which was probably due to over estimation of the ssDNA region because of the presence of local double-stranded regions generated by the secondary structure, since the Keff values were expressed in nucleotide equivalents. These results supported our conclusion that the predominant requirement for stimulation of ATPase is nonstructural ssDNA.
Evaluation of ATPase activity of PIF1C, a mutant lacking the PINT domain
We noted that the Keff value for ssDNA of PIF1 was in line with one report for Sc Pif1 ( 5 ), but significantly lower than that for human PIF1 with a N-terminal truncated form ( 4 ). We consider that the difference could be attributed to the missing function of the PINT domain. To test this possibility, we examined ATPase activity of PIF1C ( Figure 2 A), and also tested PIF1N as a control. First, we confirmed no ATPase activity of PIF1N ( Figure 2 A), as expected from the lack of any ATPase motifs. When M13 mp7 ssDNA was titrated in the reaction with PIF1C, we detected similar levels of maximum ATPase activity at higher concentrations of ssDNA ( Figure 2 A), suggesting that the helicase domain is sufficient to express ATPase activity. However, the required concentrations of ssDNA to achieve maximal activity were significantly increased ( Figure 2 A inset). The calculated Keff value, 8.7 μM (in nucleotides) for M13 mp7 ssDNAs with PIF1C was more than 20 times higher than that for PIF1 ( Table 2 ). This property was also demonstrated by determining Keff values for different oligonucleotides ( Table 2 ), calculated from data from ssDNA titration experiments ( Supplementary Figure S3B ). These results suggested that a defect in PIF1C could be in the binding to ssDNA, and the PINT domain could help interaction between the helicase domain and ssDNA.
To determine preferential nucleotide compositions for PIF1C stimulation of ATPase activity, it was measured using the same set of the oligonucleotides in the reactions with nearly saturated concentrations of the oligonucleotides (150 μM in nucleotides) ( Supplementary Figure S3B ). The results demonstrated that the preference of PIF1C was essentially identical to that of PIF1 ( Figure 2 C and D), suggesting that it is intrinsic to the helicase domain, the PINT domain not affecting this property.
For further comparison between PIF1 and PIF1C, we determined their Km values for ATP in the reactions with saturating concentrations of M13 mp18 ssDNA, 3.8 and 150 μM (in nucleotides), respectively. The Km values for ATP of both proteins were both 0.17 mM and essentially identical to the previously reported value of 0.2 mM for N-terminal-truncated human PIF1 ( 4 ). This result suggested again that the helicase domain is sufficient to express ssDNA-dependent ATPase activity.
Characterization of ssDNA binding activities of PIF1 and PIF1C
It seemed that the apparent affinity to ssDNA for the helicase domain (PIF1C) was much lower than that for PIF1 in ATPase reactions. We considered that the defect might be attributed to binding ability to ssDNA. To measure DNA binding of PIF1 and PIF1C directly, we performed electrophoretic mobility shift assays (EMSA) using a oligonucleotide, d(AC)60 ( Table 1 ) as a model substrate. It has been reported that this assay detects DNA–protein complexes in yeast and human PIF1 in an ATP-independent manner ( 7 , 9 ). Binding reactions were carried out under the optimal ATPase assay conditions, then products were loaded on gels as described in the Materials and methods section. The titration experiment displayed PIF1–DNA complexes, which were increased depending on the concentration of PIF1 ( Figure 3 A). The apparent Kd , which is approximately equal to the protein concentration at which half the free DNA has become bound ( 35 ), was determined to be about 3 nM. This was in good agreement with the Keff value of d(AC)60 for ATPase activity, because the 0.13 μM (in nucleotides) ( Table 2 ) corrected for the concentration of the 60-mer oligonucleotide became 2 nM.
EMSA of PIF1 and PIF1C for binding to ssDNA. ( A ) EMSA of ssDNA binding to PIF1 and PIF1C under standard reaction conditions. The 5′- 32 P labeled oligonucleotide, d(AC)60, was incubated with the indicated concentrations of PIF1. Arrowheads indicate the positions of free DNA. The quantified data are shown graphically. ( B ) EMSA in the absence of ATP. Experiments were performed as described in (A). The quantified data are shown graphically. ( C ) EMSA in the absence of MgCl 2 . Experiments were performed as described in (A). The quantified data are shown graphically together with data from (A) and (B). The errors in the experiments were <10%.
EMSA of PIF1 and PIF1C for binding to ssDNA. ( A ) EMSA of ssDNA binding to PIF1 and PIF1C under standard reaction conditions. The 5′- 32 P labeled oligonucleotide, d(AC)60, was incubated with the indicated concentrations of PIF1. Arrowheads indicate the positions of free DNA. The quantified data are shown graphically. ( B ) EMSA in the absence of ATP. Experiments were performed as described in (A). The quantified data are shown graphically. ( C ) EMSA in the absence of MgCl 2 . Experiments were performed as described in (A). The quantified data are shown graphically together with data from (A) and (B). The errors in the experiments were <10%.
When PIF1C was tested, we could detect PIF1C–DNA complexes, but the affinity seemed much lower than that with PIF1, demonstrating a defect in ssDNA binding activity. When the point at which half the free DNA has become bound was extrapolated from the binding curve, the apparent Kd value was estimated at about 10 nM. However, the Kd value was not directly correlated with the Keff value, since it was still 6 times lower than the apparent Keff value of 60 nM (in oligonucleotides), with the Keff value of 3.7 μM (in nucleotides) ( Table 2 ), corrected for the concentration of the 60-mer oligonucleotide.
Since it was very likely that ATP could influence the binding reaction, we performed the same experiments without ATP. With the full length PIF1, the titration curve was not affected ( Figure 3 B). In contrast that for PIF1C in the absence of ATP was significantly changed. The apparent affinity was increased to about 0.3 nM ( Figure 3 B). To determine whether the effect of ATP is caused by binding or hydrolysis, titration experiments were carried out omitting MgCl 2 to prevent hydrolysis of ATP ( Figure 3 C). Under this condition, the ATPase activity was under the background level (data not shown), suggesting that, even though trace amounts of Mg ions were present as contaminants of chemicals, the contribution was negligible. The result clearly demonstrated the binding curves to be identical to that under standard assay conditions containing ATP and MgCl 2 ( Figure 3 C), suggesting that ATP binding itself affected the interaction between ssDNA and the helicase domain.
ssDNA binding activity of the PIF1N
Our results suggested that the PINT domain plays a role in modulating the ssDNA binding activity of the helicase domain. Therefore, we carried out the same binding assays with PIF1N ( Figure 4 A). The assays detected PIF1N–DNA complexes, and the titration curves were not affected by ATP and MgCl 2 ( Figure 4 A), with an apparent Kd value of about 10 nM. These results suggested that PIF1N itself possesses ATP-independent ssDNA binding activity.
EMSA of PIF1N for binding to ssDNA. ( A ) Titration of PIF1N under different conditions. The 5′- 32 P labeled oligonucleotide, d(AC)60, was incubated with the shown concentrations of PIF1N, omitting ATP or MgCl 2 as indicated. Arrowheads indicate the positions of free DNA. The quantified data are shown graphically. ( B ) Binding ability of PIF1N to different sizes of oligonucleotides. Experiments were performed using indicated oligonucleotides as substrates under standard reaction conditions in the absence of ATP. The quantified data are shown graphically. The errors in the experiments were <10%.
EMSA of PIF1N for binding to ssDNA. ( A ) Titration of PIF1N under different conditions. The 5′- 32 P labeled oligonucleotide, d(AC)60, was incubated with the shown concentrations of PIF1N, omitting ATP or MgCl 2 as indicated. Arrowheads indicate the positions of free DNA. The quantified data are shown graphically. ( B ) Binding ability of PIF1N to different sizes of oligonucleotides. Experiments were performed using indicated oligonucleotides as substrates under standard reaction conditions in the absence of ATP. The quantified data are shown graphically. The errors in the experiments were <10%.
In contrast to the binding reactions with PIF1 and PIF1C shown in Figure 3 , further shifts of the mobility of the complexes with PIF1N in higher concentrations were observed ( Figure 4 A). This could be due to more than one protein molecule binding to the 60-mer oligonucleotide. To determine the minimal ssDNA size for binding of PIF1N, six different oligonucleotides in lengths ranging from 55 to 15 bases were subjected to binding assays. When the 55-mer was used as the substrate, the result was essentially identical to that for a 60-mer. However, further reduction of the length to 45-mer and 35-mer decreased the extent of multiple binding. With the 30-mer oligonucleotide, but not the 20-mer and 15-mer, a distinct band of the complex was observed. These results demonstrated that the minimal ssDNA size could be between 30 and 20 bases.
DNA strand annealing activity of PIF1, residing in the PINT domain
During helicase assays, we unexpectedly found that PIF1 possessed robust annealing activity. For detailed analysis, a forked substrate (1F:2L) ( Figure 5 A) was denatured by heating at 100°C for 5 min, and then incubated with PIF1 under the conditions for the standard ATPase assay omitting ATP to avoid unwinding reactions. Titration of PIF1 showed about 50% of ssDNA could be annealed within 10 min when 46 nM PIF1 was present ( Figure 5 B). To determine the region of PIF1 responsible for the annealing activity, PIF1C and PIF1N were tested for their ability to promote the reaction. The results demonstrated that PIF1N, but not PIF1C, efficiently annealed ssDNA, with activity only 3-fold lower than that of PIF1 when 1F and 2L substrates were tested ( Figure 5 B). This result indicated that the annealing activity of PIF1 resides in the PINT domain.
DNA strand annealing activity. ( A ) Schematic structures of the DNA substrates. The asterisks indicate 32 P-labeled 5′-phosphate. (B) Annealing activities of PIF1, PIF1C and PIF1N. The fork-structural partial duplex DNA substrate, 1F:2L (0.35 nM), was boiled for 5 min, then incubated with increasing levels of PIF1, PIF1C and PIF1N as indicated under standard reaction conditions omitting ATP at 30°C for 10 min. The quantified data are shown graphically. ( C ) Titration of ATP on annealing reactions mediated by PIF1 and PIF1N. The fork-structural partial duplex DNA substrate, 3F:4L (0.35 nM), was boiled for 5 min, then incubated with PIF1 (33 nM) or PIF1N (195 nM) and increasing levels of ATP under standard reaction conditions omitting MgCl 2 at 30°C for 10 min. The quantified data are shown graphically. ( D ) Time course of annealing reactions mediated by PIF1 under standard reaction conditions omitting MgCl 2 . Heat denatured substrate DNA, 1F:2L (0.35 nM), was incubated with PIF1 (46 nM). The quantified data are shown graphically. ( E ) Time course of annealing reactions mediated by PIF1N under standard reaction conditions omitting MgCl 2 . Heat denatured substrate DNA, 1F:2L (0.35 nM), was incubated with PIF1N (195 nM). The quantified data are shown graphically. ( F ) Titration of ATP on annealing reactions mediated by PIF1 and PIF1N under standard reaction conditions. A complete double stranded oligonucleotide, 3S:4S (0.35 nM) was boiled for 5 min, then incubated with PIF1 (30 nM) or PIF1N (195 nM) and increasing levels of ATP under standard reaction conditions at 30°C for 10 min. The quantified data are shown graphically. The errors in the experiments were <10%.
DNA strand annealing activity. ( A ) Schematic structures of the DNA substrates. The asterisks indicate 32 P-labeled 5′-phosphate. (B) Annealing activities of PIF1, PIF1C and PIF1N. The fork-structural partial duplex DNA substrate, 1F:2L (0.35 nM), was boiled for 5 min, then incubated with increasing levels of PIF1, PIF1C and PIF1N as indicated under standard reaction conditions omitting ATP at 30°C for 10 min. The quantified data are shown graphically. ( C ) Titration of ATP on annealing reactions mediated by PIF1 and PIF1N. The fork-structural partial duplex DNA substrate, 3F:4L (0.35 nM), was boiled for 5 min, then incubated with PIF1 (33 nM) or PIF1N (195 nM) and increasing levels of ATP under standard reaction conditions omitting MgCl 2 at 30°C for 10 min. The quantified data are shown graphically. ( D ) Time course of annealing reactions mediated by PIF1 under standard reaction conditions omitting MgCl 2 . Heat denatured substrate DNA, 1F:2L (0.35 nM), was incubated with PIF1 (46 nM). The quantified data are shown graphically. ( E ) Time course of annealing reactions mediated by PIF1N under standard reaction conditions omitting MgCl 2 . Heat denatured substrate DNA, 1F:2L (0.35 nM), was incubated with PIF1N (195 nM). The quantified data are shown graphically. ( F ) Titration of ATP on annealing reactions mediated by PIF1 and PIF1N under standard reaction conditions. A complete double stranded oligonucleotide, 3S:4S (0.35 nM) was boiled for 5 min, then incubated with PIF1 (30 nM) or PIF1N (195 nM) and increasing levels of ATP under standard reaction conditions at 30°C for 10 min. The quantified data are shown graphically. The errors in the experiments were <10%.
Since ATP is essential for helicase activity, we tested effects of ATP on the annealing reaction. To avoid the effect of unwinding, titration of ATP was carried out in the absence of MgCl 2 . The result demonstrated both PIF1 and PIF1N to be inhibited by ATP ( Figure 5 C). Time course experiments demonstrated that, in the absence of ATP, PIF1 annealed up to 70% of ssDNA, whereas, in the presence of the optimal concentration of ATP (2 mM) for helicase activity, the annealed fraction reached only 20% ( Figure 5 D). The properties of PIF1 were same as those of PIF1N, 2 mM ATP also inhibiting annealing activity from 50% to 17% ( Figure 5 E). Importantly, the ATP-titration curve and time course with PIF1 were essentially identical to those for PIF1N ( Figure 5 C–E), suggesting that the properties are intrinsic to the PINT domain.
Further confirming this conclusion, we used completely complementary strands as a substrate. The annealing product has no ssDNA region, preventing product unwinding under the standard reaction conditions containing ATP and MgCl 2 . Titration curves of ATP for PIF1 and PIF1N demonstrated again inhibitory effects, although significant fractions of the substrates were spontaneously annealed under the assay conditions ( Figure 5 F).
Effects of RPA on unwinding and annealing reactions of PIF1
Elucidation of whether PIF1 has the potential to promote unwinding and annealing reactions on RPA-coated substrates is valuable for understanding cellular functions of PIF1. To do this, first, we determined the optimal concentration of RPA for the substrate DNA binding by EMSA. The substrate, 3F:4L was used for the binding reactions at the final concentration of 0.35 nM. The results of titration of RPA are shown in Figure 6 A. When the concentration of RPA was increased to 0.9 nM, one molecule of RPA bound to at least one strand of the forked substrate. At >1.8 nM, each strand of the substrate was probably occupied by one RPA molecule ( Figure 6 A). The levels of apparent affinity of RPA for ssDNA were in good agreement with previous reports ( 38 ).
Effects of RPA on the unwinding and annealing reactions. (A) EMSA of RPA binding. The fork-structural partial duplex DNA substrate, 3F:4L (0.35 nM) was incubated with RPA at indicated concentrations on ice for 10 min under the standard reaction conditions. ( B ) Unwinding assay in the presence of RPA. The fork-structural partial duplex DNA substrate, 3F:4L (0.35 nM), was incubated with RPA at indicated concentrations on ice for 10 min under standard reaction conditions. Then, PIF1 (16 nM) (left panel) and buffer (right panel) were introduced and incubated at 30°C for 10 min. Reaction products were separated on a 15–25% polyacrylamide gel. ( C ) The quantified data of unwinding reactions in (B) are shown graphically with the RPA binding curve in (A). ( D ) EMSA of RPA binding. The fork-structural partial duplex DNA substrate, 3F:4L (0.35 nM), was boiled for 5 min, then incubated with RPA at indicated concentrations on ice for 10 min under the standard reaction conditions omitting ATP. ( E ) Annealing assay in the presence of RPA. The fork-structural partial duplex DNA substrate, 3F:4L (0.35 nM) ( Figure 5 A), was boiled for 5 min, then incubated with RPA at indicated concentrations on ice for 10 min under standard reaction conditions omitting ATP. Then PIF1 (30 nM) was introduced and incubated at 30°C for 10 min. Reaction products were separated on a 15–25% polyacrylamide gel. ( F ) The quantified data of annealing reactions in (E) are shown graphically with the RPA binding curve in (D). The errors in the experiments were <10%.
Effects of RPA on the unwinding and annealing reactions. (A) EMSA of RPA binding. The fork-structural partial duplex DNA substrate, 3F:4L (0.35 nM) was incubated with RPA at indicated concentrations on ice for 10 min under the standard reaction conditions. ( B ) Unwinding assay in the presence of RPA. The fork-structural partial duplex DNA substrate, 3F:4L (0.35 nM), was incubated with RPA at indicated concentrations on ice for 10 min under standard reaction conditions. Then, PIF1 (16 nM) (left panel) and buffer (right panel) were introduced and incubated at 30°C for 10 min. Reaction products were separated on a 15–25% polyacrylamide gel. ( C ) The quantified data of unwinding reactions in (B) are shown graphically with the RPA binding curve in (A). ( D ) EMSA of RPA binding. The fork-structural partial duplex DNA substrate, 3F:4L (0.35 nM), was boiled for 5 min, then incubated with RPA at indicated concentrations on ice for 10 min under the standard reaction conditions omitting ATP. ( E ) Annealing assay in the presence of RPA. The fork-structural partial duplex DNA substrate, 3F:4L (0.35 nM) ( Figure 5 A), was boiled for 5 min, then incubated with RPA at indicated concentrations on ice for 10 min under standard reaction conditions omitting ATP. Then PIF1 (30 nM) was introduced and incubated at 30°C for 10 min. Reaction products were separated on a 15–25% polyacrylamide gel. ( F ) The quantified data of annealing reactions in (E) are shown graphically with the RPA binding curve in (D). The errors in the experiments were <10%.
Then we examined unwinding activity of PIF1. RPA was mixed with the forked substrate, 3F:4L, under the standard reaction conditions on ice, then PIF1 was introduced and incubation was performed at 30°C for 10 min. We found that RPA did not affect unwinding reactions at low concentrations <0.4 nM ( Figure 6 B, left panel), in which majority of the substrate was RPA-free ( Figure 6 A). When the concentration of RPA reached 0.9 nM, at which almost all the oligonucleotides were occupied with RPA ( Figure 6 A), severe inhibition was observed ( Figure 6 B, left panel). At much higher concentrations of RPA, we observed unwinding products ( Figure 6 B, left panel). However, a control experiment without PIF1 revealed that the products detected at higher concentrations >1.8 nM of RPA were PIF1 independent ( Figure 6 B, right panel). The quantified results shown in Figure 6 C demonstrate no difference in the two reactions at higher concentrations of RPA (>0.9 nM). These results suggest that RPA does not enhance, but rather inhibits, the helicase activity of PIF1.
To examine annealing activity in the presence of RPA, we also determined the optimal concentration of RPA for ssDNA binding. In the reactions, we used the same substrate as for the unwinding assay (3F:4L) at the final concentration of 0.35 nM, but after denaturation by heating. The assay detected only complexes with the labeled oligonucleotide 3F (55-mer), although another fragment, 4L (51-mer), was present. The results of titration of RPA are shown in Figure 6 D. When the concentration of RPA was increased to 0.9 nM, almost all the oligonucleotides 3F and probably 4L were occupied by at least one molecule of RPA. At >1.8 nM, the oligonucleotides were occupied by two molecules of RPA ( Figure 6 D).
Next, we carried out annealing assays under the same conditions. RPA was mixed with heat denatured substrate (3F and 4L) on ice, then PIF1 was introduced with incubation at 30°C for 10 min. The result clearly demonstrated an inhibitory effect of RPA ( Figure 6 E). The quantified results for inhibition of annealing reactions well correlated inversely with the amount of RPA binding ( Figure 6 F).
DISCUSSION
In this article, we document for the first time the biochemical properties of full-length human PIF1 together with those of truncated forms consisting of individual domains. We could establish intrinsic properties of the helicase domain and functional roles of the PINT domain.
Biochemical analysis of yeast PIF1 homologs has demonstrated that they preferentially unwind forked substrates ( 1 , 5 , 7 ). We showed that the property is conserved in human PIF1. However, we found that forked substrates were not optimal with respect to stimulation of ATPase activity. Rather nonstructural ssDNA greatly stimulated ATPase activity. The finding that the Keff value for nonstructural ssDNA was lower than for other DNA molecules, including forked structures, suggested preferential binding of PIF1 to ssDNA. From these results, we suggest that PIF1 needs an ssDNA region for loading and a forked structure for entrance to the double strand region by translocation.
We present several lines of evidence that the enzymatic characters of PIF1C reflect the intrinsic properties of the helicase domain. First, PIF1C expressed ATPase activity to the level equivalent to full-length PIF1 (about 1000 min −1 ) and also equivalent to that for yeast homologs ( 5 , 7 ). Second, full-length PIF1 and PIF1C both showed a similar preference for poly(purine–pyrimidine) and polypyrimidine, but not polypurine, for stimulation of ATPase. This property is also conserved in yeast homologs ( 5 , 7 ). Third, the Km values for ATP of full-length PIF1 and PIF1C were essentially identical. This is also in agreement with a previous report for N-terminal truncated PIF1, purified as a C-terminus GST-fusion protein ( 4 ). These results suggest that the determined properties of PIF1C are intrinsic to the helicase domain, and could also exclude the possibility that the his-tag at the N-terminal of PIF1C interferes with activities of the helicase domain.
Interestingly, we noted a significant difference between PIF1 and PIF1C with regard to the required concentration of ssDNA for stimulation of ATPase activity. PIF1C needed a 20 times higher concentration and also exhibited much lower unwinding activity. The results suggest that the defects could be attributed to missing functions of the PINT domain. The difference in the Keff values could be due to lower binding affinity of the helicase domain for ssDNA. Consequently, we demonstrated ssDNA binding of PIF1 directly by EMSA. The apparent Kd value, 3 nM was in good agreement with the Keff value of 2 nM when expressed with reference to the oligonucleotide concentration, suggesting that this assay well reflected the functional interaction between ssDNA and PIF1. In this assay, as expected, we demonstrated lower affinity of PIF1C to ssDNA. We suggested that the defect in PIF1C could be due, at least in part, to lower binding affinity for ssDNA, and the PINT domain plays a role for increasing this affinity of the helicase domain. Notably, the Kd value (10 nM) of PIF1C determined by EMSA was still 6 times lower than the Keff value (60 nM in oligonucleotides) for the ATPase assay. We consider the following possible explanations for the discrepancy. We found that PIF1C exhibited a much higher affinity for ssDNA without binding of ATP. With the EMSA, an ATP-free fraction of PIF1C could exist, even in the presence of ATP. Therefore, the results could be an overestimation, due to high affinity binding of the ATP-free fraction of PIF1C. Alternatively, the PINT domain could possess another function for enhancing activity of the helicase domain by modulating the mode of interaction with ssDNA. The higher affinity of PIF1C for ssDNA without binding of ATP could be intrinsic to the helicase domain. The results suggested that ATP modulated the binding mode with ssDNA. The higher affinity to ssDNA before binding of ATP is reduced by binding of ATP. The alteration must be alternatively repeated during turnover of reactions of ATPase. With the full-length PIF1, such alteration due to ATP binding was not detected, suggesting that the PINT domain somehow could suppress the alteration during turnover of ATPase reactions. In this study, we demonstrated that the PINT domain also possesses ssDNA binding activity. We suggest that its enhancement of the activities of the helicase domain is due, at least in part, to this ssDNA binding activity.
We unexpectedly found the PINT domain to further possess ssDNA annealing activity. Among the proteins handled in this study having ssDNA binding activity, including PIF1C, PIF1N and RPA, we could detect annealing activity only with PIF1N. We consider that the annealing activity could be mediated by ssDNA binding, although not attributable to the general effect of high affinity ssDNA binding. Annealing activity has been reported to be associated with the RecQ family helicase in general ( 39–45 ). While the RecQ family is distinct from PIF1 family, annealing activity shares similar properties in common. It is located outside of the conserved helicase domains ( 39 , 40 ), is inhibited by RPA, and is ATP-independent or rather inhibited by ATP ( 39–42 , 44 ). We demonstrated that inhibition by ATP is not a consequence of the unwinding reaction, suggesting that it is an intrinsic property of the PINT domain.
At the present time, the biological significance of our findings cannot be readily assessed. Importantly, we showed that RPA inhibited unwinding and annealing reactions, suggesting that these functions of PIF1 might be restricted under particular situations in DNA metabolism. There is a marked difference from RecQ helicases, whose unwinding activity is proficient on RPA-coated ssDNA and stimulated by RPA, although annealing activity is suppressed ( 39–41 , 46–48 ). Notably, unwinding activity of Sc Pif1 was stimulated by RPA ( 1 ), but that of the fission yeast homolog, Pfh1, was inhibited ( 7 ) like human PIF1, suggesting that the outcomes in cells would differ. This could be related to the fact that budding yeast has another member of the PIF1 superfamily, Rrm3, but human and fission yeast have only one. Further analysis of the precise cellular roles of PIF1 should shed light on functions in maintenance of genetic stability.
FUNDING
Grants-in-Aid from the Ministry of Education, Culture, Sports, Science and Technology of Japan (to Y.M. and K.K.); 21st Century Center of Excellence Program from the Ministry of Education, Culture, Sports, Science and Technology of Japan (to K.K.); Health and Labour Science Research Grants (to K.K.); Grants-in-Aid for Cancer Research from the Ministry of Health, Labour and Welfare (to K.K.). Funding to pay the Open Access publication charges for this articles was provided by Grant-in-Aid from the Ministry of Education, Culture, Sports, Science and Technology of Japan (to Y.M.).
Conflict of interest statement . None declared.
ACKNOWLEDGEMENTS
We thank Dr Marc S. Wold (University of Iowa College of Medicine, Iowa City, Iowa, USA), Dr Hisaji Maki (Nara Institute of Science and Technology, Nara, Japan) and Dr Tadashi Shimamoto (Hiroshima University, Hiroshima, Japan) for, respectively, providing an RPA-expression plasmid, the M13 mp7 plasmid and the E. coli strain to produce ss M13 DNA. Several cloning vectors were obtained from National BioResource Project (NIG, Japan). We are grateful to Kumiko Mizuno, Tomoka Nakashima, Masako Okii, Fumie Okubo, Kazumi Shimamoto, Hatsue Wakayama and Mai Yoshida for their laboratory assistance.
Comments