Sequence Effect on the Formation of DNA Minidumbbells - The

Oct 19, 2017 - These non-B structures have been of great interest as they were demonstrated to participate in various cellular processes such as DNA r...
0 downloads 6 Views 1MB Size
Article pubs.acs.org/JPCB

Cite This: J. Phys. Chem. B XXXX, XXX, XXX-XXX

Sequence Effect on the Formation of DNA Minidumbbells Yuan Liu and Sik Lok Lam* Department of Chemistry, The Chinese University of Hong Kong, Shatin, New Territories Hong Kong S Supporting Information *

ABSTRACT: The DNA minidumbbell (MDB) is a recently identified non-B structure. The reported MDBs contain two TTTA, CCTG, or CTTG type II loops. At present, the knowledge and understanding of the sequence criteria for MDB formation are still limited. In this study, we performed a systematic high-resolution nuclear magnetic resonance (NMR) and native gel study to investigate the effect of sequence variations in tandem repeats on the formation of MDBs. Our NMR results reveal the importance of hydrogen bonds, base−base stacking, and hydrophobic interactions from each of the participating residues. We conclude that in the MDBs formed by tandem repeats, C− G loop-closing base pairs are more stabilizing than T−A loop-closing base pairs, and thymine residues in both the second and third loop positions are more stabilizing than cytosine residues. The results from this study enrich our knowledge on the sequence criteria for the formation of MDBs, paving a path for better exploring their potential roles in biological systems and DNA nanotechnology.



INTRODUCTION

It has long been known that B-DNA is the most common structure adopted by DNA molecules under physiological conditions.1 In addition to B-DNA, many alternative structures such as hairpins,2−4 dumbbells,5−7 triplexes,8−10 and quadruplexes11−13 have been identified in the past few decades. These non-B structures have been of great interest as they were demonstrated to participate in various cellular processes such as DNA replication, transcription, and repair.14 For example, the non-B structures can transiently exist on the nascent strand during DNA replication, providing potential pathways for the occurrence of expansion mutations.15,16 Besides, the noncytotoxic nature of DNAs makes themselves useful in constructing nanomaterials for biomedical applications.17−19 Non-B DNAs have been widely introduced into nanomaterials as they have unique structures, and can undergo structural changes in response to various environmental conditions such as changes in ionic strengths and pH values.17−20 Recently, high-resolution nuclear magnetic resonance (NMR) spectroscopic studies have revealed a new form of non-B DNA structure, namely, the minidumbbell (MDB), which comprises two type II loops (Figure 1).21−24 Type II loop is one of the two well-defined folding geometries of DNA tetraloops in which the first (L1) and fourth (L4) loop residues form a loop-closing base pair, the second loop residue (L2) sits in the minor groove, and the third loop residue (L3) stacks on L1-L4.25,26 At present, the type II loop sequences that bring about the formation of MDBs include TTTA,24 CCTG,23 and CTTG.21 Extensive intraloop and interloop interactions have been shown to be indispensable for the formation of these MDBs.21,22 Specifically, the loop-closing base pairs from the two loops (L1-L4 and L1′-L4′) stack on each other, and the second loop residues (L2 and L2′) both sit in the minor groove, interacting favorably with each other via stacking in the © XXXX American Chemical Society

Figure 1. A schematic representation of the reported MDB structures comprising two type II loops. In these loops, the second loop residues (L2 and L2′) fold into the minor groove, and interact with each other through forming a hydrogen-bonded mispair (left) or base−base stacking (right). The L3 and L3′ residues stack on the L1−L4 and L1′−L4′ loop-closing base pairs, respectively.

TTTA MDB22 or forming a T•T/C•C mispair with hydrogen bonds in the CTTG/CCTG MDB.21,22 In all these reported MDBs, the minor groove L2 and L2′ residues can also form hydrogen bonds with the loop-closing base pairs and/or the phosphodiester backbone from the same or opposite loop. In addition, the third loop residues (L3 and L3′) stack on the loop-closing base pairs, providing additional stabilizations to the MDBs. To date, all reported type II loop-forming sequences are pyrimidine-rich, including CCCG,26 CCTG,27 CTCG,27 CTTG,28 TTTA,29 and TTTG.30 They all start with a pyrimidine L1 which allows the folding of a L2 more easily into the minor groove. The more favorable folding of a pyrimidine L2 than a purine L2 into the minor groove is probably due to its smaller ring size and the presence of stabilizing interactions such as hydrogen bonds or perpendicReceived: September 7, 2017 Revised: October 19, 2017 Published: October 19, 2017 A

DOI: 10.1021/acs.jpcb.7b08904 J. Phys. Chem. B XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry B ular base−base interactions with the loop-closing base pair and the adjacent base pair.26,27,29,30 Although these pyrimidine-rich sequences are type II loop-forming sequences, their capabilities in constituting an MDB structure may differ significantly as the formation of MDBs requires both intraloop and interloop interactions.21,22 In order to better understand the sequence criteria for the formation of MDBs, we conducted a systematic study to investigate the effect of sequence variations in tandem repeats on the formation of MDBs. By comparing the solution structural behaviors of these tandem repeat sequences, we obtained deeper insights into the underlying chemical forces that account for the observed differences in the formation of MDBs.

75 ms. Data sets of 4096 × 512 and 4096 × 256 were collected for all 2D NOESY and TOCSY experiments, respectively. 2D 1 H−13C heteronuclear multiple-bond correlation (HMBC) experiments were conducted with a data size of 2048 × 200 and an evolution time of 65 ms for the long-range couplings.33 The 13C spectral width was set to 60 ppm and the carrier frequency was centered at 140 ppm. 1H−31P heteronuclear single-quantum coherence (HSQC) spectroscopy was conducted with a data size of 4096 × 200. The 31P spectral width was set to 9 ppm, and the carrier frequency was centered at −3.8 ppm. 13C and 31P chemical shifts were indirectly referenced to DSS signal using the derived nucleus-specific ratios of 0.251449530 and 0.404808636, respectively.34

EXPERIMENTAL METHODS Sample Design. As all reported MDBs contain either T−A or C−G loop-closing base pairs, we only include tandem repeats that can potentially bring about the formation of T−A or C−G loop-closing base pairs in this study. According to the folding geometry of MDBs (Figure 1), all tandem repeats of 5′CYYG CYYG-3′ and 5′-TYYA TYYA-3′ (where Y is a pyrimidine C or T) are potential MDB-forming sequences as they have the capabilities to form two CYYG or TYYA type II loops, respectively. For simplicity, these two sets of sequences were named as “(CYYG)2” and “(TYYA)2”. The former set includes “(CTTG)2”, “(CTCG)2”, “(CCTG)2” and “(CCCG)2” and the latter set includes “(TTTA)2”, “(TTCA)2”, “(TCTA)2” and “(TCCA)2”. Since the solution structural behaviors of the sequences (CTTG)2, (CCTG)2, and (TTTA)2 have been reported previously,21−24 we will focus our NMR investigations on the remaining five sequences in this study. Sample Preparation. DNA samples were synthesized using an Applied Biosystems model 394 DNA synthesizer and purified using denaturing polyacrylamide gel electrophoresis and diethylaminoethyl Sephacel anion exchange column chromatography. Sample quantities were determined by measuring the ultraviolet absorbance at 260 nm. NMR samples were prepared to contain 0.5 mM DNA in a 500 μL buffer solution with 10 mM sodium phosphate (pH 7.0), and 0.02 mM 2,2-dimethyl-2-silapentane-5-sulfonic acid (DSS). Native Gel Assay. 25% native polyacrylamide gels were prepared and the electrophoresis experiment was conducted in a pH 7.0 buffer containing 9 mM piperazine-N,N′-bis(2ethanesulfonic acid) (PIPES), 20 mM bis(2-hydroxyethyl)amino−tris(hydroxymethyl)methane (Bis−Tris) and 1 mM ethylenediaminetetraacetic acid (EDTA). The temperature of the buffer solution was maintained at ∼10 °C during electrophoresis. The loading samples were prepared to be 0.5 mM DNA in a buffer solution containing 10 mM sodium phosphate (pH 7.0), and 20% sucrose (w/v). DNA bands were visualized by poststaining the gels with stains-all solution. NMR Experiments. All NMR experiments were performed on a Bruker AV-500 and/or AV-700 NMR spectrometer(s). For studying the NMR signals from labile protons, the samples were prepared in a 90% H2O/10% D2O buffer solution. Onedimensional (1D) imino proton spectra and two-dimensional (2D) nuclear Overhauser effect spectroscopy (NOESY) were acquired using the water suppression by excitation sculpture31 or jump-return pulse sequence.32 For studying the nonlabile proton signals, the solvent was exchanged to a 99.96% D2O solution. 2D NOESY experiments were performed with a mixing time of 300, 600, and 800 ms. 2D total correlation spectroscopy (TOCSY) was performed with a mixing time of

RESULTS AND DISCUSSION Native Gel Assay. Native gel assay was first performed to determine the oligomeric states of all (CYYG)2 and (TYYA)2 sequences. The relative migration rates of DNA oligonucleotides in native gels also provide information on their secondary structures, which depend on the effective charge and frictional coefficient of molecules.35,36 If the secondary structure is relatively unstable, it will be partially unfolded and the electrophoretic mobility will be the weighted average between those of the secondary structure and random coil.35 Figure 2





Figure 2. Native gel results of (CYYG)2 and (TYYA)2 sequences. A self-complementary 8-bp duplex d(CGCTAGCG)2 was used as a reference with 150 mM sodium chloride added to increase the duplex stability. Lanes 1−8: (CTTG)2, (CCTG)2, (CTCG)2, (CCCG)2, (TTTA)2, (TTCA)2, (TCTA)2, and (TCCA)2.

shows the native gel results of (CYYG)2 and (TYYA)2 sequences at ∼10 °C. All (CYYG)2 and (TYYA)2 sequences migrated faster than the reference 8-bp duplex, suggesting they behaved as a monomolecular species. Among all eight sequences, (CTTG)2 in lane 1 migrated the fastest. The sequences (CCTG)2 in lane 2 and (TTTA)2 in lane 5 show medium migration rates. The remaining five sequences, including (CTCG) 2 , (CCCG) 2 , (TTCA) 2 , (TCTA)2, and (TCCA)2, show slower migration rates. The DNAs with sequence (CTTG)2, (CCTG)2, and (TTTA)2 have been shown to adopt MDBs with the reported melting temperature (Tm) values of ∼35, 22, and 18 °C, respectively.21,23,24 Since these MDBs are compactly folded, they are expected to migrate faster than random coils in native gels. The relative migration rates suggest that the MDB population of the sequence (CTTG)2 is larger than those of (CCTG)2 and (TTTA)2 at ∼10 °C, agreeing with the estimations from their reported Tm values. For the sequences (CTCG)2, (CCCG)2, (TTCA)2, (TCTA)2, and (TCCA)2, the slower migration rates suggest that they have smaller MDB populations than those of the three reported MDBs, and thus having lower Tm values. Resonance Assignments of (CYYG)2 and (TYYA)2 Sequences. As the solution structural behaviors of the sequences (CTTG)2, (CCTG)2, and (TTTA)2 have been investigated,21−24 NMR experiments were mainly focused on the remaining five (CYYG)2 and (TYYA)2 sequences. B

DOI: 10.1021/acs.jpcb.7b08904 J. Phys. Chem. B XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry B

downfield shifted whereas C1 and C5 H6 would be both upfield shifted. As suggested by the 1H chemical shift changes (Figure S11c), instead of an MDB, this sequence probably partially folded into a hairpin containing a GCCC loop and a 2nt 5′-CC overhang at lower temperatures (Figure S11d). Solution Structural Behaviors of (TYYA)2 Sequences. NMR structural investigations were also performed on the sequences (TTCA)2, (TCTA)2, and (TCCA)2. Figure 4a shows

Sequential resonance assignments were made from the fingerprint regions of 2D NOESY spectra using standard methods (Figures S1−S5).37 For the sequence (TTCA)2, the adenine H2 assignments were confirmed by 1H−13C HMBC experiments (Figure S6). The backbone 31P assignments were achieved by first assigning the H3′ protons in TOCSY spectra, then correlating these protons with the 31P signal of the following residue in the 1H−31P HSQC spectra (Figures S7− S9). The variable temperature 1D 1H and 31P spectral results are summarized in Figures S10−S14. Solution Structural Behaviors of (CYYG)2 Sequences. The characteristic 31P chemical shifts of the CTTG and CCTG MDBs are shown in Figure 3a and b, respectively, including two

Figure 4. Solution structural behaviors and 31P spectra of (a) the TTTA MDB reference, and the sequences (b) (TTCA)2, (c) (TCTA)2, and (d) (TCCA)2. The spectra shown here were acquired at 10 °C. Figure 3. Solution structural behaviors and 31P spectra of the (a) CTTG and (b) CCTG MDB references and the sequences (c) (CTCG)2 and (d) (CCCG)2. The spectra shown here were acquired at 10 °C.

the characteristic 31P spectrum of the reference TTTA MDB at 10 °C. For the sequence (TTCA)2, the 31P spectra show two slightly downfield shifted T2 and T6 31P signals at lower temperatures (Figure 4b and Figure S12a, left), revealing a tendency that the sequence can fold into an MDB with a relatively low thermodynamic stability. This tendency was further supported by the upfield shifted T1 and T5 H7 signals, and downfield shifted T2 and T6 H7 signals at lower temperatures (Figure S12a, right), which agree with the characteristic 31P and H7 signals of two TTCA type II loops. To further demonstrate the potential of MDB formation, 1D NOE and 2D NOESY spectra were also acquired at 0 °C. Despite the low thermodynamic stability and population, NMR evidence of the MDB with two TTCA type II loops (Figure 5a) were still observed, including (i) the 3′-5′ terminal NOEs of A8 H8/H2-T1 H7 (Figure 5b) and A8 H1′-T1 H6 (Figure S3), (ii) the 1D NOEs between A4/A8 H2 and T1/T5 imino (Figure 5c), suggesting there are two T−A Watson−Crick loop-closing base pairs, (iii) the NOEs between T2/T6 H7/ H2′/H2″ and A4/A8 H2 (Figure 5d), suggesting T2 and T6 are in the minor groove, and (iv) the NOEs of C3/C7 H5-T1/ T5 H6 and C3/C7 H5-A4/A8 H2 (Figure 5b), suggesting C3 and C7 stack on the loop-closing base pairs. For the two minor groove residues, T2 shows stronger NOEs to the loop-closing base pairs than T6 (Figure 5d). This suggests that T2 and T6 stack with each other with T2 positioned between T6 and the loop-closing base pairs. For the sequences (TCTA)2 and (TCCA)2, they were found to be mostly unstructured at 10 °C as evidenced by their 31P chemical shifts which fall within the range of random coils (∼−3.8 to −4.4 ppm)38,39 (Figure 4c and d). All the H7 chemical shifts were also found to fall into the range of random

unusually downfield (L2 and L2′) and two unusually upfield (L3 and L3′) chemical shifts. For the sequence (CTCG)2, all 31 P chemical shifts were found to locate within the range of random coils (∼−3.8 to −4.4 ppm)38,39 (Figure 3c), suggesting the sequence is unstructured. This structural finding agrees well with its relatively slower migration rate in the native gel (Figure 2). Upon lowering the temperature to ∼0 °C, peak broadenings were observed in both the 1H and 31P spectra (Figure S10a and b), revealing (CTCG)2 might form some secondary structures. To explore whether (CTCG)2 has the potential to fold into an MDB, we extracted the 1H chemical shifts from its variable temperature 1D spectra. Yet the 1H chemical shift profiles do not show any trend of MDB formation (Figure S10c). If there were the formation of two CTCG type II loops, T2 and T6 H7 signals would be both downfield shifted whereas C1 and C5 H6 would be both upfield shifted. Possibly, instead of an MDB, (CTCG)2 prefers to fold into an alternative hairpin conformer with a GCTC loop and a 2-nt 5′-CT overhang at lower temperatures (Figure S10d). The sequence (CCCG)2 was found to behave similar to (CTCG)2. At 10 °C, the sequence does not have a well-defined structure as suggested by their 31P chemical shifts, which fall within the range of random coils (Figure 3d). As suggested by the 1H and 31P peak broadenings at lower temperatures (Figure S11a and b), this sequence also shows a tendency to adopt some secondary structures. If there were the formation of two CCCG type II loops, C2 and C6 H6 signals would be both C

DOI: 10.1021/acs.jpcb.7b08904 J. Phys. Chem. B XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry B

and (TYYA)2 sequences. To rationalize the observed differences in the MDB formation propensities of these tandem repeats, we systematically analyzed the sequence effect from (i) the loop-closing base pairs, (ii) L2 residues, and (iii) L3 residues. Based on the reported three-dimensional solution structures of the CCTG and TTTA MDBs,22 we also constructed putative MDB models for these tandem repeats in order to obtain structural insights into the underlying chemical forces which govern the MDB stabilities. To understand the impacts of C−G and T−A loop-closing base pairs, we compared the relative thermodynamic stabilities and the solution structural findings of (i) the CTTG and TTTA MDBs, and (ii) the CCTG and putative TCTA MDBs. The Tm of the CTTG MDB was found to be ∼17 °C higher than that of the TTTA MDB (Table 1). For the sequence (TCTA)2, it was found to be unstructured at 0 °C whereas the sequence (CCTG)2 was known to form an MDB with a Tm of ∼22 °C. In both cases, C−G loop-closing base pairs were found to be more stabilizing than T−A loop-closing base pairs in the formation of MDBs. Thermodynamics studies have shown that C−G base pairs are generally more stabilizing than T−A base pairs.40,41 In B-DNA, a GC/CG stack is more stable than an AT/TA stack by ∼1.4 kcal·mol−1.40 In addition to the stability differences in these stacks, the reverse wobble T•T mispair in the minor groove of the CTTG MDB can form up to four hydrogen bonds with the C−G loop-closing base pairs while the T-T stack in the TTTA MDB can at most form one hydrogen bond with the T−A loop-closing base pairs (Figure S15).21 This difference possibly also makes C−G loop-closing base pairs more stabilizing than the T−A loop-closing base pairs in the formation of MDBs. For the C•C mispair in the minor groove of the CCTG MDB, it can further stabilize the MDB by forming up to three hydrogen bonds with the C−G loop-closing base pairs (Figure S16a). In the putative TCTA MDB, neither the potential C•C mispair nor C−C stack in the minor groove can form any hydrogen bond with the T−A loop-closing base pairs (Figure S16b). Therefore, the intrinsically higher stability of C−G base pairs and their higher capabilities to form hydrogen bonds with the minor groove L2/L2′ residues account for the larger stabilizing effect of C−G than T−A loop-closing base pairs on the formation of MDBs. A Thymine L2/L2′ Is More Stabilizing Than a Cytosine L2/L2′. To reveal the observed differences in the minor groove thymine versus cytosine L2/L2′ residues, we compared the solution structural behaviors between the sequences (i) (CTTG)2 and (CCTG)2 and the sequences (ii) (TTTA)2 and (TCTA)2. For the MDBs formed by the sequences (CTTG)2 and (CCTG)2, the Tm of the CTTG MDB was found to be ∼13 °C higher than that of the CCTG MDB.21 This was mainly attributed to the methyl groups of T2 and T6 involved in the hydrophobic interactions with the sugars of the C−G loop-closing base pairs in the CTTG MDB, which are absent in the CCTG MDB.21 Moreover, the reverse wobble T•T mispair in the minor groove of the CTTG MDB can also form more hydrogen bonds with the loop-closing base pairs than the C•C mispair in the CCTG MDB (Figure S15a and S16a), making thymine L2/L2′ more stabilizing than cytosine L2/L2′. For the sequences (TTTA)2 and (TCTA)2, the former forms an MDB with a Tm of ∼18 °C while the latter remains unstructured at 0 °C. This also suggests that a thymine L2/L2′ is more stabilizing than a cytosine L2/L2′ in the formation of

Figure 5. (a) Schematic representation of the MDB formed by the sequence (TTCA)2 with Watson−Crick T−A loop-closing base pairs (top). The potential hydrogen bond donors and acceptors located in the major and minor grooves of T−A base pairs are shown (bottom). (b) The 3′−5′ terminal NOEs (black dotted lines) support the folding geometry of the TTCA MDB whereas the C3−A4 and C7−A8 NOEs (red dotted line) supports C3 and C7 stack on T1−A4 and T5−A8 base pairs, respectively. The NOESY was acquired with a mixing time of 300 ms. (c) 1D NOE difference spectra by selectively irradiating T1/T5 imino at ∼12.5 ppm (red arrow) with different irradiation times of 0.3, 1, and 2 s. The appearance of A4 and A8 H2 signals supports the presence of T1-A4 and T5-A8 Watson−Crick base pairs. The 1H reference spectrum of the sequence (TTCA)2 using jumpreturn pulse sequence was shown at the bottom. (d) The NOEs between T2/T6 and A4/H8 H2 suggest T2 and T6 are in the minor groove. The NOESY was acquired with a mixing time of 600 ms. All spectra were acquired at 0 °C.

coils (∼1.7 to 1.9 ppm)38,39 (Figure S13a and S14a) and show almost no change over the entire temperature range from 0 to 75 °C (Figure S13b and S14b). C−G Loop-Closing Base Pairs Are More Stabilizing than T−A Loop-Closing Base Pairs. Table 1 shows a summary of the solution structural behaviors of all (CYYG)2 Table 1. Solution Structural Behaviors of (CYYG)2 and (TYYA)2 Sequences name (CTTG)2 (CCTG)2 (CTCG)2 (CCCG)2 (TTTA)2 (TTCA)2 (TCTA)2 (TCCA)2

sequence

structural behaviors at 10 °C

Group 1: (CYYG)2 5′-CTTG CTTG-3′ MDB 5′-CCTG CCTG-3′ predominantly MDB 5′-CTCG CTCG-3′ unstructured 5′-CCCG CCCG-3′ unstructured Group 2: (TYYA)2 5′-TTTA TTTA-3′ predominantly MDB 5′-TTCA TTCA-3′ unstructured 5′-TCTA TCTA-3′ unstructured 5′-TCCA TCCA-3′ unstructured

Tm (°C)a ∼35 ∼22 b b ∼18 b b b

a

The Tm values were reported in refs 21, 23, and 24. bThe Tm was not determined as the sequence did not fully fold into a well-defined structure at 0 °C. D

DOI: 10.1021/acs.jpcb.7b08904 J. Phys. Chem. B XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry B MDB, primarily due to the presence of hydrophobic interactions between the methyl groups of T2/T6 and the sugars of the loop-closing base pairs in the TTTA MDB. Besides, the T−T stack in the TTTA MDB can also form one hydrogen bond with the T−A loop-closing base pairs (Figure S15b).21 However, as discussed earlier, neither the C•C mispair nor C−C stack in the minor groove can form any hydrogen bond with the T−A loop-closing base pairs in the putative TCTA MDB model (Figure S16b). This also possibly contributes to the observation that a thymine L2/L2′ is more stabilizing than a cytosine L2/L2′ in the formation of MDBs. A Thymine L3/L3′ Is Also More Stabilizing Than a Cytosine L3/L3′. To show the relative contributions of thymine and cytosine L3/L3′ in the formation of MDBs, we compared the solution structural behaviors between the sequences (i) (CTTG)2 and (CTCG)2, (ii) (CCTG)2 and (CCCG)2, and (iii) (TTTA)2 and (TTCA)2. For the sequences (CTTG)2 and (CTCG)2, the former was shown to fold into the CTTG MDB with a Tm of ∼35 °C whereas the latter was found to be unstructured at ∼10 °C (Table 1), suggesting a thymine L3/L3′ is more stabilizing than a cytosine L3/L3′ in the formation of MDB. Similar findings were also observed between the sequences (CCTG)2 and (CCCG)2, in which the former was found to fold into the CCTG MDB with a Tm of ∼22 °C whereas the latter remained unstructured at ∼10 °C. Such observation was also shown in the case between the sequences (TTTA)2 and (TTCA)2 in which the TTCA MDB was found to be much less stable than the TTTA MDB. As both thymine and cytosine are pyrimidines, the gain in stability through the stacking of L3/L3′ residues on the loop-closing base pairs is expected to be similar. Therefore, the major reason making thymine more stabilizing than cytosine is attributed to the hydrophobic interactions between the methyl groups of thymine residues and the sugars of their two preceding residues in the formation of type II loops in the MDBs.21,22 Besides, in the solution structures of the CCTG and TTTA MDBs, the thymine L3/L3′ can form one to two hydrogen bonds with the C−G/T−A loop-closing base pairs (Figure S17a). Yet, in the putative CTCG, CCCG, and TTCA MDBs, the cytosine L3/ L3′ can form at most one hydrogen bond with the loop-closing base pairs (Figure S17b). This also possibly makes them less stabilizing than thymine in the formation of MDBs. Stabilizing Forces in MDBs. From studying the sequence effect on the solution structural behaviors of the two sets of tandem repeat sequences, we concluded that a change in the type II loop-forming sequence from CTTG, CCTG or TTTA to other sequences results in a weakening of the MDB structures. Our NMR findings reveal the importance of intraloop and interloop interactions for the formation of MDBs. These include hydrogen bonds, base−base stacking, and hydrophobic interactions. From the major groove view (Figure 6a), the fundamental stabilizing forces come from (i) hydrogen bonds between L1 and L4 or L1 and L4′, and (ii) base−base stacking between (a) L1−L4 and L1′−L4′, (b) L3 and L1−L4, and (c) L3′ and L1′−L4′. If the third loop residues (L3/L3′) become a thymine, there will be additional hydrophobic stabilizations. From the minor groove view, if the second loop residues form a mispair (Figure 6b, left), hydrogen bonds (i) in the mispair, and (ii) between the mispair and loop-closing base pairs will contribute to the MDB stability. If the second loop residues stack with each other (Figure 6b, right), both base− base stacking and also hydrogen bonds between L2/L2′ and the backbone phosphates will contribute to the stability. In both

Figure 6. Overview of all possible stabilizing forces in MDBs: (a) from the major groove view and (b) from the minor groove view with L2 and L2′ forming a mispair (left), or stacking with each other (right). There are additional hydrophobic interactions (blue arrows) stabilizing the MDBs when the second or third loop residues are thymine.

cases, there will be additional hydrophobic interactions if the second loop residues (L2/L2′) become a thymine. In short, the observed sequence effect on the formation of MDBs is well rationalized with the different types of stabilizing forces in this study, justifying the sequence (CTTG)2 forms the most stable MDB at neutral pH condition among all 8-nt native DNA tandem repeat sequences.



CONCLUSIONS On the basis of the systematic comparisons of the solution structural behaviors of all (CYYG)2 and (TYYA)2 sequences, the results obtained in this study reveal that (i) C−G loopclosing base pairs are more stabilizing than T−A loop-closing base pairs, (ii) thymine L2 residues are more stabilizing than cytosine L2 residues, and (iii) thymine L3 residues are also more stabilizing than cytosine L3 residues in the formation of MDBs. In addition, the better understanding of the sequence effect expands our knowledge of the sequence criteria in the formation of MDBs, allowing further explorations of the potential biological roles of MDBs and their applications in DNA nanotechnology.



ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jpcb.7b08904. 1 H and 31P resonance assignments, chemical shift profiles, alternative conformers, and models showing the stabilizing interactions (PDF)



AUTHOR INFORMATION

Corresponding Author

*Telephone: +852 3943 8126. Fax: +852 2603 5057. E-mail: [email protected]. ORCID

Sik Lok Lam: 0000-0001-5368-706X E

DOI: 10.1021/acs.jpcb.7b08904 J. Phys. Chem. B XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry B Notes

(21) Liu, Y.; Guo, P.; Lam, S. L. Formation of a DNA mini-dumbbell with a quasi-type II loop. J. Phys. Chem. B 2017, 121, 2554−2560. (22) Guo, P.; Lam, S. L. Minidumbbell: a new form of native DNA structure. J. Am. Chem. Soc. 2016, 138, 12534−12540. (23) Guo, P.; Lam, S. L. New insights into the genetic instability in CCTG repeats. FEBS Lett. 2015, 589, 3058−3063. (24) Guo, P.; Lam, S. L. Unusual structures of TTTA repeats in icaC gene of Staphylococcus aureus. FEBS Lett. 2015, 589, 1296−1300. (25) Hilbers, C. W.; Heus, H. A.; van Dongen, M. J. P.; Wijmenga, S. S. The Hairpin Elements of Nucleic Acid Structure: DNA and RNA Folding. Nucleic Acids Mol. Biol. 1994, 8, 56−104. (26) van Dongen, M. J.; Wijmenga, S. S.; van der Marel, G. A.; van Boom, J. H.; Hilbers, C. W. The transition from a neutral-pH double helix to a low-pH triple helix induces a conformational switch in the CCCG tetraloop closing a Watson-Crick stem. J. Mol. Biol. 1996, 263, 715−729. (27) Ippel, H. H.; van den Elst, H.; van der Marel, G. A.; van Boom, J. H.; Altona, C. Structural similarities and differences between H1and H2-family DNA minihairpin loops: NMR studies of octameric minihairpins. Biopolymers 1998, 46, 375−393. (28) Wu, B.; Girard, F.; van Buuren, B.; Schleucher, J.; Tessari, M.; Wijmenga, S. Global structure of a DNA three-way junction by solution NMR: towards prediction of 3H fold. Nucleic Acids Res. 2004, 32, 3228−3239. (29) Blommers, M. J.; van de Ven, F. J.; van der Marel, G. A.; van Boom, J. H.; Hilbers, C. W. The three-dimensional structure of a DNA hairpin in solution. Two-dimensional NMR studies and structural analysis of d(ATCCTATTTATAGGAT). Eur. J. Biochem. 1991, 201, 33−51. (30) Chou, S. H.; Chin, K. H.; Chen, C. W. Enhanced loop DNA folding induced by thymine-CH3 group contact and perpendicular guanine-thymine interaction. J. Biomol. NMR 2001, 19, 33−48. (31) Stott, K.; Stonehouse, J.; Keeler, J.; Hwang, T.-L.; Shaka, A. J. Excitation sculpting in high-resolution nuclear magnetic resonance spectroscopy: application to selective NOE experiments. J. Am. Chem. Soc. 1995, 117, 4199−4200. (32) Plateau, P.; Gueron, M. Exchangeable proton NMR without base-line distorsion, using new strong-pulse sequences. J. Am. Chem. Soc. 1982, 104, 7310−7311. (33) van Dongen, M. J.; Wijmenga, S. S.; Eritja, R.; Azorin, F.; Hilbers, C. W. Through-bond correlation of adenine H2 and H8 protons in unlabeled DNA fragments by HMBC spectroscopy. J. Biomol. NMR 1996, 8, 207−212. (34) Markley, J. L.; Bax, A.; Arata, Y.; Hilbers, C. W.; Kaptein, R.; Sykes, B. D.; Wright, P. E.; Wüthrich, K. Recommendations for the presentation of NMR structures of proteins and nucleic acids (IUPAC Recommendations 1998). Pure Appl. Chem. 1998, 70, 117−142. (35) Stellwagen, E.; Abdulla, A.; Dong, Q.; Stellwagen, N. C. Electrophoretic mobility is a reporter of hairpin structure in singlestranded DNA oligomers. Biochemistry 2007, 46, 10931−10941. (36) Viovy, J.-L. Electrophoresis of DNA and other polyelectrolytes: physical mechanisms. Rev. Mod. Phys. 2000, 72, 813. (37) Wijmenga, S. S.; van Buuren, B. N. M. The use of NMR methods for conformational studies of nucleic acids. Prog. Nucl. Magn. Reson. Spectrosc. 1998, 32, 287−387. (38) Lam, S. L.; Chi, L. M. Use of chemical shifts for structural studies of nucleic acids. Prog. Nucl. Magn. Reson. Spectrosc. 2010, 56, 289−310. (39) Ho, C. N.; Lam, S. L. Random coil phosphorus chemical shift of deoxyribonucleic acids. J. Magn. Reson. 2004, 171, 193−200. (40) SantaLucia, J., Jr.; Hicks, D. The thermodynamics of DNA structural motifs. Annu. Rev. Biophys. Biomol. Struct. 2004, 33, 415− 440. (41) Yakovchuk, P.; Protozanova, E.; Frank-Kamenetskii, M. D. Basestacking and base-pairing contributions into thermal stability of the DNA double helix. Nucleic Acids Res. 2006, 34, 564−574.

The authors declare no competing financial interest.



ACKNOWLEDGMENTS We thank Dr. Pei Guo for her valuable comments and suggestions on this manuscript. The work described in this paper was supported by General Research Fund (CUHK14302915) and Special Equipment Grant (SEG/ CUHK09) from the Research Grants Council of the Hong Kong Special Administrative Region, and Direct Grant (4053194) from the Faculty of Science of The Chinese University of Hong Kong.



REFERENCES

(1) Watson, J. D.; Crick, F. H. C. Molecular structure of nucleic acids: a structure for deoxyribose nucleic acid. Nature 1953, 171, 737− 738. (2) Bikard, D.; Loot, C.; Baharoglu, Z.; Mazel, D. Folded DNA in action: hairpin formation and biological functions in prokaryotes. Microbiol. Mol. Biol. Rev. 2010, 74, 570−588. (3) Chi, L. M.; Lam, S. L. Structural roles of CTG repeats in slippage expansion during DNA replication. Nucleic Acids Res. 2005, 33, 1604− 1617. (4) Varani, G. Exceptionally stable nucleic acid hairpins. Annu. Rev. Biophys. Biomol. Struct. 1995, 24, 379−404. (5) Lam, S. L.; Wu, F.; Yang, H.; Chi, L. M. The origin of genetic instability in CCTG repeats. Nucleic Acids Res. 2011, 39, 6260−6268. (6) Rinkel, L. J.; Tinoco, I., Jr. A proton NMR study of a DNA dumbbell structure with hairpin loops of only two nucleotides: d(CACGTGTGTGCGTGCA). Nucleic Acids Res. 1991, 19, 3695− 3700. (7) Erie, D.; Sinha, N.; Olson, W.; Jones, R.; Breslauer, K. A dumbbell-shaped, double-hairpin structure of DNA: a thermodynamic investigation. Biochemistry 1987, 26, 7150−7159. (8) Rajeswari, M. R. DNA triplex structures in neurodegenerative disorder, Friedreich’s ataxia. J. Biosci. 2012, 37, 519−532. (9) Vetcher, A. A.; Napierala, M.; Iyer, R. R.; Chastain, P. D.; Griffith, J. D.; Wells, R. D. Sticky DNA, a long GAA.GAA.TTC triplex that is formed intramolecularly, in the sequence of intron 1 of the frataxin gene. J. Biol. Chem. 2002, 277, 39217−39227. (10) Frank-Kamenetskii, M. D.; Mirkin, S. M. Triplex DNA structures. Annu. Rev. Biochem. 1995, 64, 65−95. (11) Rhodes, D.; Lipps, H. J. G-quadruplexes and their regulatory roles in biology. Nucleic Acids Res. 2015, 43, 8627−8637. (12) Burge, S.; Parkinson, G. N.; Hazel, P.; Todd, A. K.; Neidle, S. Quadruplex DNA: sequence, topology and structure. Nucleic Acids Res. 2006, 34, 5402−5415. (13) Fojtík, P.; Vorlícková, M. The fragile X chromosome (GCC) repeat folds into a DNA tetraplex at neutral pH. Nucleic Acids Res. 2001, 29, 4684−4690. (14) Mirkin, S. M. Expandable DNA repeats and human disease. Nature 2007, 447, 932−940. (15) Wells, R. D. Non-B DNA conformations, mutagenesis and disease. Trends Biochem. Sci. 2007, 32, 271−278. (16) Bacolla, A.; Wells, R. D. Non-B DNA conformations as determinants of mutagenesis and human disease. Mol. Carcinog. 2009, 48, 273−285. (17) Seeman, N. C. An overview of structural DNA nanotechnology. Mol. Biotechnol. 2007, 37, 246−257. (18) Pinheiro, A. V.; Han, D.; Shih, W. M.; Yan, H. Challenges and opportunities for structural DNA nanotechnology. Nat. Nanotechnol. 2011, 6, 763−772. (19) Choi, J.; Majima, T. Conformational changes of non-B DNA. Chem. Soc. Rev. 2011, 40, 5893−5909. (20) Guo, P.; Lam, S. L. An Extraordinarily Stable DNA Minidumbbell. J. Phys. Chem. Lett. 2017, 8, 3478−3481. F

DOI: 10.1021/acs.jpcb.7b08904 J. Phys. Chem. B XXXX, XXX, XXX−XXX