Non-Watson–Crick interactions between PNA and DNA inhibit the ATPase activity of bacteriophage T4 Dda helicase

Abstract

Peptide nucleic acid (PNA) is a DNA mimic in which the nucleobases are linked by an N-(2-aminoethyl) glycine backbone. Here we report that PNA can interact with single-stranded DNA (ssDNA) in a non-sequence-specific fashion. We observed that a 15mer PNA inhibited the ssDNA-stimulated ATPase activity of a bacteriophage T4 helicase, Dda. Surprisingly, when a fluorescein-labeled 15mer PNA was used in binding studies no interaction was observed between PNA and Dda. However, fluorescence polarization did reveal non-sequence-specific interactions between PNA and ssDNA. Thus, the inhibition of ATPase activity of Dda appears to result from depletion of the available ssDNA due to non-Watson–Crick binding of PNA to ssDNA. Inhibition of the ssDNA-stimulated ATPase activity was observed for several PNAs of varying length and sequence. To study the basis for this phenomenon, we examined self-aggregation by PNAs. The 15mer PNA readily self-aggregates to the point of precipitation. Since PNAs are hydrophobic, they aggregate more than DNA or RNA, making the study of this phenomenon essential for understanding the properties of PNA. Non-sequence-specific interactions between PNA and ssDNA were observed at moderate concentrations of PNA, suggesting that such interactions should be considered for antisense and antigene applications.

INTRODUCTION

Peptide nucleic acid (PNA) is a synthetic mimic of DNA that serves as an agent for antisense and antigene applications (1,2). Recently, PNAs targeted to the terminus of the 5′-UTR of luciferase mRNA have been shown to be highly effective antisense agents in cell culture (2). Another application has been for the inhibition of human telomerase activity in which PNA was targeted to the RNA component of the enzyme (3,4). Additionally, convenient methods for the cellular uptake of PNA have been demonstrated (5,6).

The chemical composition of PNA contains elements of DNA and peptides (7). The N-(2-aminoethyl) glycine backbone of PNA is drastically different to the sugar–phosphate backbone of DNA (Fig. 1). The NMR structure of a PNA–DNA hybrid reveals a helix that has components of A- and B-form DNA (8). PNA contains the same nucleobases as DNA and they are oriented to generate Watson–Crick interactions with a complementary strand of DNA, RNA or PNA. PNA heteroduplexes have greater thermal stability relative to nucleic acid duplexes presumably due to the lack of charge on the ‘peptide-like’ backbone of PNA. The increase in thermal stability of PNA heteroduplexes appears to be independent of the ionic strength of the solvent (9).

Figure 1.

Structures of DNA and PNA.

PNA has recently been applied to investigate the substrate specificity of helicases. Helicases are enzymes involved in various aspects of nucleic acid metabolism including replication, recombination, repair and transport (10). Helicases are proposed to transduce the chemical energy from ATP hydrolysis into mechanical energy necessary for unwinding double-stranded nucleic acid. Dda is a helicase from bacteriophage T4 that has been implicated in T4 DNA replication and recombination (11,12). The ATPase activity of most helicases, including Dda, is stimulated upon binding of single-stranded nucleic acid, which is proposed to result in translocation of the helicase on single-stranded DNA (ssDNA) (13,14). PNA was substituted for one of the strands of a duplex DNA substrate in order to investigate whether specific, enzyme–DNA interactions were necessary for unwinding to occur (15). Unwinding of the DNA–PNA heteroduplex was as efficient as the unmodified duplex DNA substrate, suggesting that the rate-limiting step for unwinding does not involve specific interactions between the enzyme and the backbone of the displaced strand of DNA. In contrast, substitution of one of the strands of a duplex DNA substrate with a strand of PNA strongly inhibited the unwinding activity of the NS3 helicase, from the hepatitis C virus, suggesting that this helicase requires specific interactions with both strands during unwinding (16). Thus, in addition to its role as an antisense or antigene agent, PNA has served as a probe for studying enzyme–nucleic acid interactions.

The absence of charge on the PNA backbone may result in the relatively low solubility of PNA, which can be enhanced by attaching a charged amino acid such as lysine (17). Accompanying the low solubility of PNA is the tendency for PNA to self-aggregate (17). The formation of these aggregates must be considered in antisense or antigene applications of PNA. In the work presented here, we show that the intrinsic nature of PNAs to self-aggregate extends to non-sequence-specific interaction with ssDNA.

MATERIALS AND METHODS

Reagents

ATP, Triton X-100, acetonitrile, Tris, MgCl₂, KCl, BME, glycerol, bromophenol blue, acrylamide and DMF were purchased from Fisher. DIPEA, TFA, pyruvate kinase/lactate dehydrogenase, NADH, PEP, xylene cyanole and salmon testes DNA (stDNA; denatured, ssDNA fragments for hybridization) were purchased from Sigma.

PNA synthesis

PNA was synthesized by a solid-phase, F-moc procedure as previously described by Goodwin et al. (18). Fluorescein labeling of PNA was performed on the resin by adding 5- (and 6-) carboxyfluorescein succinimidyl ester (Molecular Probes) at a ratio of 3 mg/µmol PNA in 30:1 DMF and DIPEA, respectively. PNA was purified on an analytical C₁₈ column (Rainin) using a linear gradient of H₂O/0.1% TFA to acetonitrile/0.1% TFA. Elution of the PNA was monitored by using a Beckman System Gold-125 solvent module equipped with a Beckman Diode Array Detector Module-168. The fluorescein-labeled PNA was separated from the unlabeled PNA by simultaneously monitoring for A_{260 nm} and A_{492 nm}. Purified PNA was dried in vacuo using a Speed Vac (Savant) and suspended in H₂O. Purified PNA was analyzed by MALDI-TOF mass spectrometry as previously described by Goodwin et al. (18). Under denaturing conditions, the nucleobases are considered completely unstacked and the extinction coefficient at 260 nm is the sum of the extinction coefficients of the individual nucleobases (17,19). We have measured the absorbance of PNA under denaturing conditions of high temperature (80°C) or in 0.2 M KOH at room temperature and found that both methods provide identical results (K. D. Raney and A. J. Tackett, unpublished data). Thus, the A₂₆₀ of an aliquot of PNA was measured in 0.2 M KOH to determine the concentration of PNA stock solutions. The sequence (5′→3′) of the 15mer PNA used in this study is NH₂-Gly-CATCATGCAGGACAG-Lys. The extinction coefficient for this sequence is 170 600 M^–1 cm^–1.

ATPase assays

ATP hydrolysis was monitored by using a coupled, spectrophotometric assay (14). The change in absorbance when NADH is oxidized to NAD⁺ was measured at 380 nm over a 40 s period by using an Ultrospec 2000 spectrophotometer (Pharmacia Biotech) equipped with SWIFT-Reaction Kinetics software (version 1.14T). Dda rapidly hydrolyzes ATP (see Fig. 6, V_max = 3676 nM s^–1). The ΔAbs for NADH oxidation can be monitored at 340 nm; however, the relatively low concentration of NADH that can be used at 340 nm is depleted quickly by the rapid ATPase activity of Dda helicase. At 380 nm, more NADH can be added to the solution because the extinction coefficient is lower at the longer wavelength, therefore more measurements can be made with the same sample (14). The temperature of the reaction was maintained at 25°C. The spectrophotometric ATPase reaction mixture consisted of 25 mM Tris–Cl pH 7.5, 10 mM MgCl₂, 150 mM KCl, 2 mM BME, 5 mM ATP, 16.6 U/ml pyruvate kinase, 31.3 U/ml lactate dehydrogenase, 4 mM PEP, 0.7 mg/ml NADH and 25 nM Dda. Dda was expressed and purified as previously described (20). ATPase assays for determining IC₅₀ values were performed in the presence of 6 µM stDNA and varying concentrations of PNA. Other ATPase assays were performed at a constant PNA concentration by titrating increasing amounts of stDNA into the reaction. The concentration of stDNA was determined in terms of 15mers, by division of the nucleotide concentration by 15, to allow direct comparison with the concentration of 15-PNA. The nucleotide (nt) concentration for stDNA was determined by using an extinction coefficient of 8250 M_(nt)^–1 cm^–1. The rates of ATP hydrolysis were determined by measuring the slope of the absorbance change at A_{380 nm} as a function of time using an extinction coefficient of 1210 M^–1 cm^–1 for NADH oxidation (14). Hydrolysis of 1 mol ATP is equivalent to the oxidation of 1 mol NADH.

Figure 6.

Inhibition of Dda ATPase activity in the presence of 15-PNA. (A) Double-reciprocal plot of stDNA-stimulated ATPase activity of 25 nM Dda in the presence of varying 15-PNA: 0 µM (closed circles), 0.5 µM (open squares), 1 µM (open diamonds), 3 µM (×), 5 µM (+) and 10 µM (open triangles) 15-PNA. The inhibition patterns were fit to a linear equation. (B) Direct plot of the 15-PNA inhibition data from (A). The data in the absence of 15-PNA (closed circles) were fit to equation 2 (V_max = 3676 ± 50 nM s^–1; K_act = 3.6 ± 0.17 µM). The data in the presence of 15-PNA were fit to equation 4. (C) Direct plot of the 10 µM 15-PNA inhibition data from (B). The solid line is the fit obtained using equation 4. The broken line is the fit to equation 2.

At a fixed concentration of stDNA, plots of ATPase rates as a function of PNA concentration were fit to equation 1 with KaleidaGraph software to determine IC₅₀ values:

% Activity = 100/{1 + ([PNA]/IC₅₀)} 1

Double-reciprocal plots of ATPase rates versus stDNA concentration at fixed concentrations of 15-PNA were fit to a linear function with KaleidaGraph software. Direct plots of ATPase rates versus stDNA concentration in the absence of inhibitor were fit to equation 2, which provides the steady-state rate for ATP hydrolysis under conditions of saturating concentration of ATP (21):

ν = V_max[D]/(K_act + [D]) 2

Equation 2 describes ssDNA-stimulated ATPase activity of a helicase where D is the concentration of stDNA, V_max is the ATPase activity when stDNA is saturating, and K_act is the concentration of stDNA required to achieve half the maximum ATPase activity (21). As described, 15-PNA interacts with stDNA, and effectively lowers the concentration of stDNA that is available for binding to the helicase. To account for the apparent depletion in stDNA concentration, equation 3 (22) was substituted into equation 2. Equation 3 describes the equilibrium binding of 15-PNA to stDNA.

3

K₀ is the equilibrium constant describing the formation of the DNA–inhibitor complex (as shown in Scheme 2), D_T is the total concentration of stDNA and D is the concentration of free stDNA in the presence of 15-PNA (I). Equation 3 accounts for the interaction between stDNA and 15-PNA (as shown in Scheme 2). The substitution of equation 3 into equation 2 gives rise to equation 4:

Scheme 2. Kinetic model for inhibition of ssDNA-stimulated ATPase activity of Dda by PNA.

4

The plots of ATPase activity as a function of stDNA concentration in the presence of PNA were fit to equation 4 to describe the substrate depletion.

Fluorescence polarization experiments

Fluorescence polarization was measured by using a Beacon Fluorescence Polarization instrument (PanVera). Fluorescence polarization buffer consisted of 25 mM Tris–Cl pH 7.5, 10 mM MgCl₂, 150 mM KCl and 2 mM BME. A fluorescein-labeled 15mer DNA (F15-DNA) containing an identical sequence to the 15-PNA was purchased from Integrated DNA Technologies, purified by preparative gel electrophoresis, and quantified by A₂₆₀ with an extinction coefficient of 183 600 M^–1 cm^–1. A series of 1 ml samples of F15-DNA (1 nM) or F15-PNA (1 nM) was prepared in fluorescence polarization buffer and titrated with varying amounts of ligand, either Dda or 15-PNA. Fluorescence polarization buffer was used as the blank and a sample with 1 nM F15-DNA or 1 nM F15-PNA was used as the zero point to monitor the change in milli-polarization units (mP). Fluorescence polarization samples were incubated for 3 min at 25°C prior to each measurement. Fluorescence polarization data of F15-DNA association with Dda was fit to a hyperbola by using KaleidaGraph software. For fluorescence polarization experiments performed in the presence of 1% Triton X-100, the non-ionic surfactant was included into the fluorescence polarization buffer prior to the incubation of PNA. The fluorescence polarization experiments at elevated temperature were performed by incubating each sample in a 60°C water bath for 3 min prior to measuring fluorescence polarization.

Gel-shift experiments

15-DNA was 5′-radiolabeled with [γ-³²P]ATP (PerkinElmer Life Sciences) using T4 polynucleotide kinase (New England BioLabs) as described previously (15,16). To create duplexes, radiolabeled 15-DNA was mixed with an equivalent of complementary 15mer DNA or complementary 15mer PNA. To analyze non-Watson–Crick interactions, 50 nM of radiolabeled 15-DNA was incubated with 10 µM 15-PNA (same sequence as 15-DNA). Each sample was mixed with non-denaturing load buffer (0.1% bromophenol blue, 0.1% xylene cyanol, 6% glycerol). Samples were resolved by electrophoresis on a 20% non-denaturing polyacrylamide gel and visualized with a phosphorimager.

RESULTS AND DISCUSSION

PNA inhibits the ATPase activity of a DNA helicase

Helicases unwind dsDNA, producing ssDNA products that can spontaneously reanneal. In order to prevent reannealing, a ssDNA trapping strand can be included in the reaction mixture. However, the ssDNA can bind to the helicase and slow observed steady-state rates of DNA unwinding (23). We have shown how PNA can replace ssDNA as the trapping strand, leading to faster observed rates of unwinding because the PNA does not sequester the helicase (24).

While evaluating the limitations of applying PNA as a trapping strand for DNA unwinding experiments, we observed inhibition of the ssDNA-stimulated ATPase activity of the Dda helicase in the presence of high concentrations of a 15mer PNA that was a different sequence to that used previously. stDNA was used in these assays to provide a random, single-strand sequence. The concentration of stDNA was determined in terms of 15mers to allow direct comparison with the 15-PNA. The assay used an enzymatic, coupling system to provide ATP regeneration (pyruvate kinase) and a mode for measuring ATP hydrolysis (lactate dehydrogenase oxidation of NADH). In the presence of saturating ATP (5 mM), Dda (25 nM) exhibited no measurable ATP hydrolysis, and increasing concentrations of PNA (0.5, 1, 3, 5 and 10 µM) do not stimulate ATP hydrolysis (15). Thus, PNA does not fulfill the chemical requirements necessary to stimulate the ATPase activity of Dda. When increasing concentrations of 15-PNA were added to a reaction mixture of 25 nM Dda plus 6 µM stDNA, the observed ATPase activity was inhibited (Fig. 2). Fitting the data in Figure 2 to equation 1 provided an IC₅₀ = 5.1 ± 0.39 µM for 15-PNA (Table 1). Attempts to observe inhibition at concentrations higher than 10 µM 15-PNA were unsuccessful due to precipitation of the PNA.

Figure 2.

Inhibition of Dda ATPase activity at increasing concentrations of 15-PNA. The ATPase activity of 25 nM Dda was stimulated with 6 µM stDNA and the inhibition of ATPase activity at increasing concentrations of 15-PNA was followed with a spectrophotometric assay. The data was fit to equation 1, which produced an IC₅₀ value of 5.1 ± 0.39 µM.

Table 1. Inhibition of Dda ATPase activity by PNAs.

PNA	Sequence	IC50 (µM)
15-PNA	5′-(N-ter)-Gly-CATCATGCAGGACAG-Lys-(C-ter)-3′	5.1 ± 0.39
15-PNA (complement)	5′-(N-ter)-Gly-CTGTCCTGCATGATG-Lys-(C-ter)-3′	38 ± 3.2
6-PNA	5′-(N-ter)-TTGGGG-(C-ter)-3′	7.9 ± 1.8
13-PNA	5′-(N-ter)-Gly–GGTTAGACAAAAA-Lys-(C-ter)-3′	23 ± 6.5

Dda does not interact with PNA

We postulated that the bacteriophage T4 helicase, Dda, might interact with PNA weakly through base-stacking interactions as observed in the crystal structure of the Escherichia coli Rep helicase bound to ssDNA (25). Such an interaction might explain the inhibition of the ATPase activity. To investigate this possibility, fluorescence polarization was used to detect potential interactions between 15-PNA and the Dda helicase. Fluorescence polarization has been used to study interactions between DNA and proteins as well as protein–protein interactions because the signal is related to the molecular size of the fluorophore-labeled molecule (26,27). When a protein binds to the fluorescently labeled molecule such as DNA, the rotational freedom of the DNA is usually reduced, leading to an increase in polarization. One advantage of using fluorescence polarization is the ability to observe formation of multi-species complexes in solution under equilibrium conditions. Fluorescein-labeled 15mer DNA (1 nM) (F15-DNA) and 1 nM fluorescein-labeled 15mer PNA (F15-PNA) were titrated separately with increasing concentrations of Dda (Fig. 3). The F15-DNA and F15-PNA were identical in sequence in order to avoid any sequence-specific effects. Dda interacted with F15-DNA with a K_D = 62 ± 3.2 nM under these conditions. However, no change in polarization was observed with F15-PNA under these conditions, suggesting that Dda does not bind to F15-PNA (Fig. 3). This suggests that inhibition of ATPase activity by the 15-PNA is not caused by a direct interaction between PNA and the Dda helicase.

Figure 3.

Fluorescence polarization (mP) of 1 nM F15-DNA (circles) and 1 nM F15-PNA (squares) with increasing concentrations of Dda. F15-DNA and Dda associated with a K_D = 62 ± 3.2 nM. F15-PNA and Dda did not interact. The data were fit to a hyperbola.

PNA forms non-Watson–Crick interactions with ssDNA

The finding that a 15mer PNA inhibits Dda ATPase activity stimulated by stDNA, but does not bind directly to Dda, suggested that PNA might interact with the stDNA resulting in reduced binding of Dda to stDNA. PNA forms Watson–Crick and Hoogsteen interactions with complementary strands of nucleic acid (1,7,9). However, the possibility of other interactions has not been fully explored. PNA is a hydrophobic species and the formation of self-aggregates has been suggested (17). PNA interactions with complementary DNA have recently been studied with fluorescence polarization (28). While complementary PNA–DNA duplexes show little change in fluorescence polarization, the addition of a high molecular weight polymer like polylysine will interact with the nucleic acid and significantly alter the rotational correlation time (28). Additionally, fluorescence polarization has been used to study molecular aggregation (29–32). Thus, this technique provides a direct method for investigating the aggregation properties of PNA.

When a strand of 1 nM F15-DNA was titrated with a complementary strand of 15mer PNA up to 500 nM, a change of only ∼30 mP was observed (Fig. 4A). This is consistent with previous reports of small changes in fluorescence polarization occurring when PNA hybridizes to complementary strands of DNA (28). The data in Figure 4A show the expected change in fluorescence polarization for normal Watson–Crick hybridization between a 15mer DNA and 15mer PNA. To investigate possible alternative interactions between the 15-PNA and a non-complementary sequence of DNA, 1 nM F15-DNA was titrated with unlabeled 15-PNA (Fig. 4B). The F15-DNA and 15-PNA are of identical, non-palindrome sequence, thus the formation of a DNA–PNA hybrid due to normal Watson–Crick interactions would not be expected. However, a change in fluorescence polarization was observed up to 10 µM 15-PNA (Fig. 4B). The addition of higher concentrations of 15-PNA was not possible due to precipitation. The large change in fluorescence polarization suggests that the interaction of 15-PNA with the non-complementary strand of DNA involved multiple strands of PNA because simple hybridization with a complementary strand leads to only small changes in polarization (Fig. 4A). These data do not eliminate formation of some Watson–Crick base pairing in the aggregate, but the 120 mP change in polarization (Fig. 4B) is likely due to a ‘non-Watson–Crick’ aggregate forming between multiple strands of PNA and the identical sequences of DNA. Due to the weak interaction between the PNA and DNA (Fig. 4B) and the low solubility of the PNA (∼10 µM), a stoichiometric analysis of the aggregate was not possible because it would require higher concentrations of PNA.

Figure 4.

Interaction of 15-PNA with DNA. (A) Fluorescence polarization of 1 nM F15-DNA titrated with a complementary 15mer PNA in binding buffer. Only a 30 mP change was observed for the Watson–Crick interaction. (B) Fluorescence polarization (mP) of 1 nM F15-DNA titrated with the identical sequence of 15-PNA in binding buffer at 25°C (closed circles), binding buffer + 1% Triton X-100 at 25°C (open squares), and binding buffer at 60°C (closed squares). Fluorescence polarization (mP) of 1 nM F24-DNA titrated with the non-complementary 15-PNA (open circles). (C) Fluorescence polarization of 1 nM F15-PNA titrated with increasing amounts of random sequence, stDNA. No interaction was observed.

To further investigate PNA aggregation, 1 nM F15-DNA was titrated with non-complementary 15-PNA in the presence of 1% Triton X-100. The non-ionic surfactant did not prevent formation of the aggregate (Fig. 4B). However, titration of 1 nM F15-DNA with 15-PNA at elevated temperature (60°C) prevented formation of the aggregate (Fig. 4B). Thus, the DNA–PNA aggregate is temperature sensitive. Additionally, a non-complementary fluorescein labeled 24mer (F24-DNA) was titrated with the 15-PNA, which resulted in a slightly larger change in fluorescence polarization than that observed with the F15-DNA (Fig. 4B). As with F15-DNA, a complex forms between the 15-PNA and an oligonucleotide that is non-complementary to the PNA.

In a separate experiment, 1 nM F15-PNA was titrated with increasing concentrations of non-complementary ssDNA, and no change in polarization was observed (Fig. 4C). Therefore, the change in polarization was observed only when the concentration of the PNA was increased rather than when DNA concentration was increased. These findings suggest that the interaction is due to excess formation of PNA aggregates at high concentrations of PNA. The PNA aggregates interact with ssDNA in a non-Watson–Crick manner. To further test this hypothesis, interaction of PNA with itself was investigated.

PNA forms self-aggregates

Non-specific interaction of PNA with DNA might be preceded by aggregation of the PNA with itself, prior to binding to DNA. To address this question, 1 nM F15-PNA was titrated with the exact same sequence of unlabeled 15-PNA (Fig. 5). Interaction between F15-PNA and 15-PNA was observed up to 10 µM 15-PNA after which precipitation occurred. The change in fluorescence polarization increased with each addition of 15-PNA (Fig. 5). Fluorescence polarization is a function of molecular weight, thus self-aggregates are forming. The formation of self-aggregates is a cumulative process that likely results in precipitation. The titration was also performed in the presence of 1% Triton X-100 and separately at 60°C. The PNA aggregate was stable in the presence of the surfactant, whereas it was unstable at elevated temperature (Fig. 5).

Figure 5.

Self-interactions of 15-PNA. Fluorescence polarization (mP) of 1 nM F15-PNA titrated with the identical sequence of 15-PNA in binding buffer at 25°C (circles), binding buffer + 1% Triton X-100 at 25°C (open squares), and binding buffer at 60°C (closed squares).

PNA inhibits Dda ATPase activity by substrate depletion

The presence of excess PNA promotes aggregate formation. These results suggest that inhibition of Dda ATPase activity by 15-PNA may be due to non-Watson–Crick interactions between PNA and ssDNA (Scheme 1), thus leading to depletion of the available ssDNA for productive binding by the helicase (Scheme 2). This proposed mechanism for inhibition of Dda ATPase activity through non-Watson–Crick interactions between 15-PNA and ssDNA was investigated directly by varying stDNA concentration at fixed concentrations of 15-PNA using the spectrophotometric ATPase assay. When stDNA was titrated into a reaction mixture containing Dda and 15-PNA, a linear correlation was observed in a double-reciprocal plot of velocity versus concentration of stDNA (Fig. 6A). Titrating stDNA at higher concentrations of 15-PNA in the reaction reduced the ATPase activity as illustrated in Figure 6A. These plots resemble competitive inhibition, however, the lines converge in the first quadrant rather than on the y-axis (Fig. 6A). The intersection of lines in the first quadrant as observed in Figure 6A is consistent with the inhibition pattern for substrate depletion (33). To ensure that the observed inhibition was not influenced by 15-PNA interacting with the enzymatic coupling system, TLC-based ATPase assays were performed in the absence of the coupling system. The amount of inhibition observed in the presence of 15-PNA was essentially identical between the spectrophotometric and TLC assays, so the observed inhibition is not associated with the coupling system (data not shown). Additionally, when using the coupled, spectrophotometric assay, complete recovery of ATPase activity could be obtained by adding additional stDNA, which overcame the DNA depletion due to binding by the PNA, thereby revealing that PNA inhibition was not a result of 15-PNA interacting with the ATPase coupling system.

Scheme 1. Model for non-sequence-specific interaction between PNA aggregate and ssDNA at relatively high concentrations of PNA.

Directly plotting the ATPase activity versus the stDNA concentration provided insight into the apparent competitive inhibition (Fig. 6B). The sigmoidal trend observed with increasing 15-PNA concentration is consistent with substrate depletion (22). The sigmoidal shape is due to inhibitor (15-PNA) depleting DNA substrate (stDNA) from the reaction. As the concentration of 15-PNA is increased from 0.5 to 10 µM, the sigmoidal shape becomes more apparent because of greater substrate depletion (Fig. 6B). At increasing concentrations of 15-PNA the sigmoidal shape should become more defined, however, 15-PNA precipitated in the reaction conditions beyond 10 µM. According to the fluorescence polarization experiments, 15-PNA interacts with ssDNA in a non-Watson–Crick manner and thereby prevents Dda from binding productively to the stDNA as described in Scheme 2, which supports the trend observed in Figure 6B.

In the absence of inhibitor (15-PNA), the rate equation derived from Scheme 2 (equation 2) can be used to fit data for ATPase activity as a function of DNA concentration (Fig. 6B, circles) (21). In the presence of PNA, the effective DNA concentration is depleted due to binding of DNA by PNA, and a sigmoidal trend in the data is apparent. The reduction in the concentration of free DNA available for binding to the helicase can be described by equation 3, where K₀ represents the equilibrium constant for the formation of the DNA–PNA complex (22). By substitution of equation 3 for the DNA concentration in equation 2, an equation (equation 4) was generated that described the DNA depletion due to binding by PNA. The 15-PNA inhibition data were fit to equation 4 (Fig. 6B). This approach is appropriate when the helicase concentration is very low relative to the DNA concentration and takes into account the depletion of DNA due to binding of PNA (Scheme 2).

Fitting the data to equation 4 produced a K₀ = 0.91 ± 1.32 µM (Fig. 6B). The 15-PNA concentrations were relatively close to the K₀ and the non-Watson–Crick interactions between the PNA and DNA were weak, therefore the K₀ value determined is only qualitative. Also, because the PNA forms aggregates of undetermined size, the effective concentration of the aggregates is not known. Higher concentrations of PNA would better define the K₀, however, precipitation occurred beyond 10 µM 15-PNA, which prevented further analysis. For illustration, the 10 µM 15-PNA inhibition data is shown alone in Figure 6C. The solid line is the fit using equation 4, and the broken line is the fit with only equation 2. The sigmoidal character of the data is defined with equation 4, which describes the loss of DNA due to substrate depletion through interactions with PNA. Thus, a model invoking DNA depletion in which interaction occurs between 15-PNA and the stDNA can describe the data for inhibiton of ATPase activity (Scheme 2).

PNA aggregates observed by DNA gel-shift analysis

15-DNA was 5′-radiolabeled to observe interactions with complementary and non-complementary PNA. Samples were analyzed by non-denaturing PAGE and visualized by phosphorimaging. When radiolabeled 15-DNA was annealed with complementary DNA or complementary PNA, duplexes were formed as expected (Fig. 7). The DNA–PNA duplex did not migrate as far as the DNA–DNA duplex presumably due to the lack of charge on the PNA backbone (15,16). When the non-complementary 15-PNA was incubated with radiolabeled 15-DNA, aggregate formation was observed (Fig. 7). The aggregate appeared as a large species that did not migrate into the gel. These experiments show formation of the aggregate between 15-PNA and an identical sequence of DNA, indicating that non-Watson–Crick interactions are most likely responsible for formation of the aggregate. Formation of the large molecular mass species that was retained in the well supports the fluorescence polarization data in Figure 4B. The aggregate was more pronounced in 10 mM HEPES pH 7.5 + 1 mM EDTA (Fig. 7A), relative to the ATPase reaction conditions (Fig. 7B). When Dda (25 nM) was added to the sample, the aggregate was somewhat resolved (Fig. 7B, lane 5). Heating the sample resolved the aggregate (Fig. 7B, lane 6). Dda has been shown to displace streptavidin from biotin-labeled oligonucleotides (13) and to unwind DNA–PNA hybrids (15). Dda binds tightly to ssDNA whereas the PNA aggregates bind weakly. Thus, the ability of Dda to resolve the DNA–PNA aggregate is consistent with the enzyme’s ability to remove many different types of molecules from ssDNA.

Figure 7.

Gel-shift analysis of the aggregate between 15-DNA and 15-PNA. (A) Aggregate formation in 10 mM HEPES (pH 7.5) + 1 mM EDTA. Phosphorimage of: lane 1, radiolabeled 15-DNA; lane 2, radiolabeled 15-DNA annealed to complementary 15mer DNA; lane 3, radiolabeled 15-DNA annealed to complementary 15mer PNA; and lane 4, radiolabeled 15-DNA plus 10 µM non-complementary 15-PNA. Each sample was resolved on a 20% polyacrylamide gel and visualized with a phosphorimager. (B) Aggregate formation in the ATPase reaction buffer. Phosphorimage of: lane 1, radiolabeled 15-DNA; lane 2, radiolabeled 15-DNA annealed to complementary 15mer DNA; lane 3, radiolabeled 15-DNA annealed to complementary 15mer PNA; lane 4, radiolabeled 15-DNA plus 10 µM non-complementary 15-PNA; lane 5, radiolabeled 15-DNA, 10 µM non-complementary 15-PNA and 25 nM Dda; lane 6, radiolabeled 15-DNA, 10 µM non-complementary 15-PNA and 25 nM Dda heated to 95°C for 10 min.

Inhibition of Dda ATPase activity by PNA is not specific

The inhibition of Dda ATPase activity was not confined to the 15-PNA. A series of PNAs varying in length and sequence were analyzed for inhibition (Table 1). Each PNA produced very similar inhibitory patterns as seen for 15-PNA in Figures 2 and 6. The apparent IC₅₀ values for each sequence of PNA were in the low micromolar range (Table 1). Variations in IC₅₀ values do not appear to be length dependent because of the difference between the two 15mers (Table 1). Base composition may play a role in the degree of inhibition, however, each PNA appears to inhibit Dda ATPase activity by substrate depletion through non-specific interactions with ssDNA. A 15mer PNA sequence that was previously used as a trapping strand for DNA unwinding experiments also inhibited the ATPase activity of Dda, however, the IC₅₀ was 10-fold higher than concentrations used in the previous experiments (24).

CONCLUSION

The data presented reveal the ability of PNA to aggregate with identical sequences of ssDNA and self-aggregate in a non-Watson–Crick manner (Scheme 1). These PNA aggregates inhibited the ATPase activity of the Dda helicase through depletion of ssDNA as described in Scheme 2. PNA did not bind measurably to Dda under the conditions described here. The inhibition is not specific to the 15-PNA because of the similar inhibition observed for the other PNA sequences as shown in Table 1. In a broad sense, these results uncover an experimentally important property of PNA: when PNA is present at relatively high concentrations, aggregation and non-specific interaction with ssDNA can occur. PNA clearly can bind with very high affinity and sequence specificity to DNA, however, for applications requiring moderate to high concentrations of PNA, the possibility of PNA aggregation and non-specific interaction between PNA and DNA should be considered.