Base pair opening kinetics study of the aegPNA:DNA hydrid duplex containing a site-specific GNA-like chiral PNA monomer.

Peptide nucleic acids (PNA) are one of the most widely used synthetic DNA mimics where the four bases are attached to a N-(2-aminoethyl)glycine (aeg) backbone instead of the negative-charged phosphate backbone in DNA. We have developed a chimeric PNA (chiPNA), in which a chiral GNA-like γ(3)T monomer is incorporated into aegPNA backbone. The base pair opening kinetics of the aegPNA:DNA and chiPNA:DNA hybrid duplexes were studied by NMR hydrogen exchange experiments. This study revealed that the aegPNA:DNA hybrid is much more stable duplex and is less dynamic compared to DNA duplex, meaning that base pairs are opened and reclosed much more slowly. The site-specific incorporation of γ(3)T monomer in the aegPNA:DNA hybrid can destabilize a specific base pair and its neighbors, maintaining the thermal stabilities and dynamic properties of other base pairs. Our hydrogen exchange study firstly revealed the unique kinetic features of base pairs in the aegPNA:DNA and chiPNA:DNA hybrids, which will provide an insight into the development of methodology for specific DNA recognition using PNA fragments.

INTRODUCTION

Peptide nucleic acids (PNA) are one of the most widely used synthetic DNA mimics where the four bases are attached to a N-(2-aminoethyl)glycine (aeg) backbone (Figure 1A) instead of the negative-charged phosphate backbone in DNA (1,2). Today, PNAs have been used for various application purposes such as: (i) potential drug candidates for gene therapy; (ii) molecular tools in biotechnology; (iii) lead compounds for the development of gene-targeted drugs; (iv) diagnostic biosensors (3). PNA is capable of sequence-specific recognition to DNA as well as RNA and forms various hybrid complexes such as PNA:PNA, PNA:DNA, PNA:RNA duplexes and PNA:PNA:PNA, PNA:DNA:PNA triplexes (4–8). UV melting studies of PNA hybrids revealed that the PNA sequences can bind to their complementary DNA, RNA and PNA strands with great affinities (5,9–11). Thermodynamic studies revealed that the hybridization of DNA with complementary PNA strand significantly increased the thermal stability of the double helix compared to DNA duplex, mainly due to the lack of interstrand electrostatic repulsion (12–14).

Figure 1.

The chemical structures of (A) aegPNA, (B) (R)-γ3T, (C) chiPNA with a (R)-γ3T. (D) Sequence contexts of the duplex samples studied.

However, drawbacks such as poor cellular uptake, ability to form hybrids with both parallel and antiparallel orientation and similiar binding affinity to DNA/RNA are known to be the subjects for improvement. Especially, the latter two may lead to poor specificity toward the target sequence, which give rise to undesirable effects both in biological and diagnostic applications. Recently, we have developed a structurally simple chiral PNA monomer with thymine base (named γ3T, Figure 1B) and prepared a chimeric PNA (named chiPNA, Figure 1C), in which γ3T monomers are incorporated into aegPNA backbone (15). This chiPNA displays excellent RNA selectivity as well as antiparallel selectivity compared to aegPNA although the melting temperatures of most chiPNA:DNA hybrids are slightly lower than the corresponding aegPNA:DNA hybrids (15). Previous hydrogen exchange study using ammonia reported that the imino protons in the aegPNA:DNA hybrid had similar base pair stabilities but very short base pair lifetimes compared to the corresponding DNA duplex (16). In order to account for the change in thermal stability and base pair dynamics of nucleic acid duplexes upon introduction of a new type of monomer in aegPNA:DNA hybrid, we performed the NMR exchange studies on the: (i) DNA:DNA duplex (DD-11); (ii) aegPNA:DNA hybrid (aegPD-11); or (iii) chiPNA:DNA hybrid (chiPD-11) which was formed by a single-stranded aegPNA containing a site-specific γ3T and its complementary DNA strand (Figure 1D). These data showed the unique kinetic features of base pairs in the aegPNA:DNA and chiPNA:DNA hybrids, which will provide an insight into the development of methodology for specific DNA recognition using PNA fragments.

MATERIALS AND METHODS

Sample preparation

Fmoc-protected aegPNA monomers were purchased from PANAGENE (Daejeon, Korea), where the exocyclic amino groups of A, G, C are Bhoc-protected. The (R)-γ3T PNA monomer was synthesized as described in our previous report (15). The aegPNA and (R)-γ3T–modified chiPNA oligomers were synthesized by using standard solid phase Fmoc-chemistry using Peptide synthesizer (ABI 433A, Applied Biosystems Inc., USA). The DNA oligomers were purchased from M-biotech Inc. (Seoul, Korea). The PNA and DNA oligomers were purified by reverse-phase HPLC and desalted using a Sephadex G-25 gel filtration column. The three duplex samples were prepared by dissolving the aegPNA, chiPNA and DNA strands with sequence of 5′-CGTGCTCGGAC-3′ (A strand in Figure 1D) and the complementary DNA strand, 5′-GTCCGAGCACG-3′ (B strand in Figure 1D) at a 1:1 stoichiometric ratio in a 90% H2O/10% D2O aqueous solution containing 10 mM TRIS-d11 (pH 8.0 at 24.2°C) and 100 mM NaCl.

NMR experiments

All NMR experiments were performed on a Varian Inova 600 MHz spectrometer (KAIST, Daejeon) using a HCN triple-resonance probe. One dimensional (1D) NMR data were processed with either the program VNMR J (Varian, Palo Alto) or FELIX2004 (Accelrys, San Diego), whereas 2D data were processed with the program NMRPIPE (17) and analyzed with the program Sparky (18). The apparent longitudinal relaxation rate constants (R1a = 1/T1a) of the imino protons were determined by semi-selective inversion recovery 1D NMR experiments. The apparent relaxation rate constant of water (R1w) was determined by a selective inversion recovery experiment, using a DANTE sequence for selective water inversion (19). R1a and R1w were determined by curve fitting of the inversion recovery data to the appropriate single-exponential function. The hydrogen exchange rate constants (kex) of the imino protons were measured by a water magnetization transfer experiment (19–22). The intensities of each imino proton were measured with 20 different delay times. The kex for the imino protons were determined by fitting the data to Equation (1): where I0 and I(t) are the peak intensities of the imino proton in the water magnetization transfer experiments at times zero and t, respectively, and R1a and R1w are the apparent longitudinal relaxation rate constants for the imino proton and water, respectively, measured in semi-selective inversion recovery 1D NMR experiments. The formalism of TRIS-catalyzed proton exchange has been previously described in detail in the previous study (19,21,22) and is briefly presented here. The exchange time for the base paired imino proton, τex (inverse of kex), is represented by: where τ0 is the base pair life time (inverse of opening rate constant, kop), ki is the rate constant for imino proton transfer, and αKop (= αkop/kcl) is the apparent equilibrium constant for base pair opening. The ki values for TRIS-catalyzed imino proton transfer are calculated by Equation (3): where kcoll is the collision rate constant for TRIS (= 1.5 × 109 s−1), and ΔpKa is the pKa difference between the imino proton (G-H1: 9.24; T-H3: 9.60) and the TRIS base (25°C, 8.15). The actual concentrations of TRIS were calculated using equation [TRIS] = [TRIS]total/(1 + 10(pa − pH)). Curve fitting the τex values of the imino protons as a function of the inverse of [TRIS] with Equation (2) gives the αKop and τ0 (= 1/kop) values.

RESULTS AND DISCUSSION

Temperature dependency of 1D imino proton spectra

Resonance assignments for the imino protons were made by the analysis of NOESY spectra. Figure 2 shows the temperature dependence of 1D imino proton spectra of DD-11, aegPD-11 and chiPD-11 duplexes. Most imino resonances are well resolved in the 1D spectra because the nucleotide sequence was designed to minimize overlapping of these resonances. All imino resonances except those of the terminal base pairs and the neighboring A10·T13 base pairs in three duplexes were intact over 25°C (Figure 2), indicating that the three duplexes form stable double helices at room temperature. Interestingly, the melting temperature of aegPD-11 was 10°C higher than those of DD-11 and chiPD-11 (Figure 2), which is consistent with the previous thermodynamic studies (12–14). In chiPD-11, the T6 and G16 imino resonances have lower melting temperature than those of aegPD-11 (Figure 2). These results clearly demonstrates that the T6·A17 and C7·G16 base pairs and the overall helical structure are destabilized upon the incorporation of γ3T at T6 position into the aegPNA strand.

Figure 2.

Temperature-dependent 1D imino proton spectra for the (A) DD-11, (B) aegPD-11, and (C) chiPD-11 duplexes in a 90% H2O/10% D2O solution containing 10 mM TRIS-d11 (pH = 8.0, 25°C) and 100 mM NaCl.

Hydrogen exchange experiments for imino protons

The hydrogen exchange rate constants (kex) determined from water magnetization transfer experiment on the imino protons for DD-11, aegPD-11 and chiPD-11 duplexes at 15°C (Figure 3) (19–23). Some imino protons show large change in peak intensities as a function of delay time after water inversion (Figure 3). For example, rapidly exchanging imino protons such as T13 in three duplexes and T6 in chiPD-11 show negative peaks at delay time of 50 ms (Figure 3), whereas the G8 resonance in DD-11, which is the slowest exchanging imino proton, remains basically unchanged up to 100 ms. The relative peak intensities, I(t)/I0, for the T6 and G16 imino proton resonances of three duplexes are plotted as a function of delay time in Figure 3D and E, respectively. These data were used to determine kex by fitting Equation (1). The kex values for individual imino protons in the DD-11, aegPD-11 and chiPD-11 duplexes determined at 15°C are shown in Table 1. Many of the imino protons exchanged too slowly to be accurately determined from this NMR hydrogen exchange method (kex ≤ 1.5 s−1) at 15°C because the peak intensities for these protons could be affected by not only solvent exchange but also direct or spin-diffused NOE interaction with water (20).

Figure 3.

1D spectra of the water magnetization transfer experiments showing the imino protons of the (A) DD-11, (B) aegPD-11 and (C) chiPD-11 duplexes in NMR buffer at 15°C. The relative peak intensities in the spectra, I(t)/I0, as a function of delay time for the (D) T6 and (E) G16 imino protons in the DD-11 (closed triangle), aegPD-11 (gray circle), and chiPD-11 (open square) duplexes at 15°C. Solid lines indicate the best fitting of these data using Equation (1).

Table 1.

The hydrogen exchange rate constants of the imino protons, kex (s−1), for the DD-11, aegPD-11 and chiPD-11 duplexes

Imino	15°Ca			25°Cb
	DD-11	aegPD-11	chiPD-11	DD-11	aegPD-11	chiPD-11
G2	1.5 ± 0.3c	7.3 ± 0.7	7.8 ± 0.8	4.1 ± 0.9c	15.9 ± 8.0	n.d.d
T3	6.6 ± 0.4	1.7 ± 0.6	1.6 ± 1.6	54.0 ± 3.1	13.6 ± 1.5	14.0 ± 2.5
G4	≤1.5	≤1.5	2.0 ± 0.8	≤1.5	≤1.5	3.0 ± 1.3
G18	1.5 ± 0.3c	≤1.5	1.6 ± 0.6	4.1 ± 0.9c	≤1.5	4.0 ± 0.9
T6	2.4 ± 0.4	≤1.5	29.8 ± 2.3	38.8 ± 1.3	1.6 ± 0.7	>400
G16	≤1.5	≤1.5	5.6 ± 0.8	≤1.5	≤1.5	53.6 ± 1.5
G8	≤1.5	≤1.5	≤1.5	≤1.5	1.6 ± 0.7	2.4 ± 1.0
G9	1.5 ± 0.3c	≤1.5	≤1.5	4.1 ± 0.9c	1.5 ± 0.6	2.0 ± 1.1
T13	121 ± 21	84.3 ± 9.2	74.1 ± 5.8	>400	>400	>400

aSample conditions: pH = 8.0, 100 mM NaCl, [TRIS]total = 10 mM, 15°C.

bSample conditions: pH = 8.0, 100 mM NaCl, [TRIS]total = 40 mM, 25°C.

cThe G2, G9 and G18 imino proton resonances overlap in the DD-11 duplex.

dThe G2 imino proton resonance shows severe line-broadened in the chiPD-11 duplex.

To further characterize the more stable regions in the three duplexes, the kex measurements were performed at 25°C. The T13 imino resonances in the three duplexes and T6 imino resonance in chiPD-11 were not observed at 25°C, which means the kex > 400 s−1 (20). Interestingly, the kex values of the T3 imino protons in aegPD-11 and chiPD-11 are 4-fold smaller than that of DD-11 at 25°C. In addition, the central T6 imino proton in aegPD-11 has kex value of 1.6 s−1, which is 24-fold smaller than that of the DD-11 duplex at 25°C. These results demonstrate that the hybridization of DNA with aegPNA strand significantly reduces the kex values for the imino protons of the A·T base pairs but has little effects on the G imino protons (Table 1). However, the T6 imino resonance in chiPD-11 was not observed, which means that its kex value at 25°C is larger than 400 s−1 (20). At 15°C, the peak intensity of T6 imino proton in chiPD-11 depends more significantly on delay time compared to DD-11 and aegPD-11 (Figure 3D), demonstrating that the specific base pair in chiPNA:DNA is more unstable than the corresponding aegPNA:DNA and DNA:DNA base pairs. This instability also affects the kex value of the neighboring base pairs. For example, the G16 imino proton in chiPD-11 has kex value of 53.6 s−1, which is 36-fold larger than the kex values of DD-11 and aegPD-11 at 25°C.

Base pair opening kinetics

The apparent equilibrium constants (αKop) for base pair opening in the DD-11, aegPD-11 and chiPD-11 duplexes can be determined by base-catalyzed imino proton exchange measurements. Ammonia is the most commonly used base catalyst in thermodynamic and kinetic study for base pair opening in the nucleic acid duplexes. However, ammonia causes severe line broadening of the imino proton resonances in very unstable nucleic acid duplexes at high temperature because of its high catalytic efficiency. For example, some imino protons in the P1 duplex of Tetrahymena ribozyme, containing a central G·U wobble pair, exhibited substantial line broadening when [NH3] >30 mM; thus, the αKop for opening of these base pairs could not be determined by ammonia catalyst (19). In the case of chiPD-11, T6 imino proton resonance were not observed even at 25°C and this instability also affects the kex value of the neighboring G16 imino proton. Therefore, TRIS base was thought to be the most suitable catalyst for the hydrogen exchange studies to examine the effect of the incorporation of a site-specific γ3T monomer into the aegPNA:DNA hybrid duplex. The kex (= 1/τex) for each imino proton in the DD-11, aegPD-11 and chiPD-11 duplexes was measured at 25°C as a function of the TRIS concentration (Figure 4). From these data, the equilibrium constants for base pair opening (αKop) and the base pair lifetimes (τ0 = 1/kop) of the three duplexes were determined by curve fitting using Equation (2) (Figure 4). In DD-11, the three G·C base pairs determined here have αKop ≤ 1 × 10−6 and τ0 > 100 ms (Table 2). These data can be used to calculate an apparent lifetime for base pair opening (ατopen = α/kcl) of 70–100 ns (using ατopen = τ0αKop). However, the A·T base pairs have much larger αKop values (T3·A20: 67 × 10−6 and T6·A17: 21 × 10−6) than the G·C base pairs and their base pair lifetimes could not be determined (τ0 < 1 ms; Table 2 and Figure 4).

Figure 4.

The hydrogen exchange times (τex) of the (A) T3, (B) G4, (C) G18, (D) T6, (E) G16 and (F) G8 imino protons of the DD-11 (closed triangle), aegPD-11 (gray circle) and chiPD-11 (open square) duplexes as a function of the inverse of TRIS concentration at 25°C. Solid lines indicate the best fitting of these data using Equation (2).

Table 2.

The equilibrium constants for base pair opening (αKop), base pair lifetimes (τ0) and apparent lifetimes for base pair opening (ατopen) of the DD-11, aegPD-11 and chiPD-11 duplexes determined by TRIS-catalyzed NMR exchange experiments at 25°C

Parameter	Base pair	DD-11	aegPD-11	chiPD-11
αKop (×10−6)	T3·A20	67 ± 3	21 ± 2	26 ± 2
	G4·C19	0.9 ± 0.1	3.5 ± 3.9	4.1 ± 1.0
	C5·G18	n.d.b	1.6 ± 0.2	3.3 ± 0.4
	T6·A17	21 ± 1	2.5 ± 0.2	n.a.c
	C7·G16	0.4 ± 0.2	0.5 ± 0.1	30 ± 3
	G8·C15	0.4 ± 0.1	1.1 ± 0.1	1.5 ± 0.1
	G9·C14	n.d.b	1.1 ± 0.1	1.4 ± 0.1
τ0 (ms)	T3·A20	<1	17 ± 2	23 ± 2
	G4·C19	120 ± 20	720 ± 88	229 ± 17
	C5·G18	n.d.b	376 ± 25	69 ± 9
	T6·A17	<1	140 ± 21	n.a.c
	C7·G16	270 ± 234	602 ± 47	<1
	G8·C15	175 ± 53	201 ± 18	68 ± 14
	G9·C14	n.d.b	171 ± 7	87 ± 14
ατopen (ns)	T3·A20	<10	350 ± 73	589 ± 79
	G4·C19	100 ± 24	2520 ± 3130	936 ± 289
	C5·G18	n.d.b	600 ± 124	232 ± 57
	T6·A17	<10	347 ± 83	n.a.c
	C7·G16	107 ± 139	325 ± 56	10 ± 29
	G8·C15	69 ± 26	225 ± 35	101 ± 28
	G9·C14	n.d.b	190 ± 15	125 ± 28

aParameters used in the calculations: kcoll = 1.5 × 109 (s-1), pKa(G-H1) = 9.24, pKa(T-H3) = 9.60, pKa(TRIS, 25°C) = 8.15; Sample conditions: pH = 8.0, 100 mM NaCl, [TRIS]total = 10–300 mM, 25°C. The actual concentrations of TRIS were calculated using equation [TRIS] = [TRIS]total/(1 + 10(pa-pH)).

bNot determined because the G18 imino proton resonance overlaps with G9 imino proton resonance.

cNo resonances.

These data were used to compare the effect of the hybridization of DNA with aegPNA on the base pair stabilities (Figure 4). In aegPD-11, as expected from the kex data, the αK values of the T3·A20 and T6·A17 base pairs are 3- and 8-fold smaller than those of DD-11, respectively (Figure 4A and D; Table 2). This means that the A·T base pairs in the aegPNA:DNA hybrid are more stable than those of the corresponding DNA duplex. However, the G4·C19 and G8·C15 base pairs in aegPD-11 have larger αK values than those of DD-11 (Figure 4B and F; Table 2). These results indicate that the PNA–DNA hybridization, even though some G·C base pairs are slightly destabilized, stabilizes the A·T base pairs and increases the overall thermal stability of the aegPNA:DNA hybrid, compared to the corresponding DNA duplex. These results were consistent with the previous hydrogen exchange study using ammonia in which the T·A base pairs in the aegPNA:DNA hybrid had 3 to 4 times smaller αKop values, but the αKop of the G·C base pairs were 2- to 3-fold larger than those of the corresponding DNA duplex at various temperatures (16).

In Figure 4, the slope of the linear correlation between τex (= 1/kex) of the imino proton and inverse of TRIS concentration is 1/(kiαKop) and the y-intercept is τ0 [see Equation (2)]. The more stable T3·A20 base pair in aegPD-11 have both a larger slope (smaller αKop) and a larger y-intercept (longer τ0) than those of DD-11 in this plot (Figure 4A), demonstrating that these base pairs in the aegPNA:DNA hybrid are opened much more slowly compared to the DNA duplex. The similar effects are also observed in the T3·A20 base pairs in aegPD-11 (Figure 4D). Surprisingly, the G4·C19 base pair of aegPD-11 has significantly a larger y-intercept (longer τ0) than that of DD-11, in spite of its instability (Figure 4B). Contrary to our results, the previous ammonia-catalyzing exchange study showed that lifetimes (τ0) for both T·A and G·C base pairs in the aegPNA:DNA hybrid were much shorter (≤1 ms even at 10°C) than those of the corresponding DNA duplex, indicating extremely fast base pair opening/closing (16). However, the previous study had been performed for relatively unstable duplexes, which contained six T·A base pairs in decamer duplex, by using strong catalyst, ammonia at pH = 8.8. These experimental conditions could make imino proton exchanges more rapid for the DNA and aegPNA:DNA hybrid duplexes. Consequently, the base pair lifetimes (τ0) may not be correctly determined by ammonia catalyst for the aegPNA:DNA hybrid, although the previous study provided the αKop value for each base pair and the important information about the base pair stabilities in the aegPNA:DNA hybrid. In our current, we could show more clearly the unique features of base pair opening in aegPNA:DNA hybrid, in which most base pairs are opened and reclosed much more slowly than those of the corresponding DNA duplex.

In chiPD-11, the αKop for the T6·A17 base pair could not be determined because of the severe line-broadening of the T6 imino protons, indicating that the specific base pair (γ3T·A17) in chiPNA:DNA base pair is much more unstable than the corresponding pairs in aegPNA:DNA and DNA:DNA duplexes. This difference in the base pair stability between aegPNA:DNA and chiPNA:DNA base pairs also affects the stabilities of the neighboring base pairs. The αKop of the C7·G16 base pair [3′-side neighbor of γ3T (T6)] in chiPD-11 is about 60-fold larger than that of aegPD-11 (Table 2). Also, its base pair lifetime could not be determined, meaning that τ0 < 1 ms. The C5·G18 base pair (the 5′-side neighbor) has only 2-fold larger αKop and a 5-fold shorter τ0 than those of aegPD-11 (Table 2), which means that the γ3T·A17 base pairing in the chiPNA:DNA hybrid leads to great destabilization and fast opening of the neighoring base pairs in an asymmetric manner. In Figure 4, the G4·C19 and G8·C15 base pairs of chiPD-11 have similar slopes but smaller y-intercepts than those of DD-11 in the plots. This means that the γ3T·A17 base pair does not affect on the thermal stabilities of next neighboring base pairs but decreases their base pair lifetimes. The T3·A20 base pair in chiPD-11 has the same αKop and τ0 values with those of aegPD-11 (Table 2), indicating that the γ3T·A17 base pair does not affect the thermal stability of this base pair as well as its base pair opening property.

SUMMARY

The aegPNA:DNA hybrid is much more stable duplex and is less dynamic compared to DNA duplex, meaning that base pairs are opened and reclosed much more slowly. The site-specific incorporation of γ3T monomer in the aegPNA:DNA hybrid can destabilize a specific base pair and its neighbors, maintaining the thermal stabilities and dynamic properties of other base pairs. Our hydrogen exchange study firstly revealed the unique kinetic features of base pairs in the aegPNA:DNA and chiPNA:DNA hybrids, which will provide an insight into the development of methodology for specific DNA recognition using PNA fragments.

FUNDING

Basic Research Promotion Fund (No. KRF-205-2005-1-C00043); National Research Foundation of Korea (NRF) grants funded by the Korean government (MEST) (No. 2011-0001318; 2010-0014199; NRF-C1ABA001-2010-0020480); Next-Generation BioGreen 21 Program, Rural Development Administration, Korea (SSAC, No. PJ008109). Funding for open access charge: National Research Foundation of Korea.

Conflict of interest statement. None declared.