Sequence-Dependent Effects of Monovalent Cations on the Structural

Subscriber access provided by Kaohsiung Medical University

B: Biophysics; Physical Chemistry of Biological Systems and Biomolecules

Sequence-Dependent Effects of Monovalent Cations on the Structural Dynamics of Trinucleotide-Repeat DNA Hairpins Marisa Lynn Mitchell, Michael Patrick Leveille, Roman Symeon Solecki, Thao Tran, and Brian Cannon J. Phys. Chem. B, Just Accepted Manuscript • DOI: 10.1021/acs.jpcb.8b07994 • Publication Date (Web): 16 Nov 2018 Downloaded from http://pubs.acs.org on November 19, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Sequence-dependent Effects of Monovalent Cations on the Structural Dynamics of Trinucleotide-Repeat DNA Hairpins Marisa L. Mitchell, Michael P. Leveille, Roman S. Solecki, Thao Tran, and Brian Cannon* AUTHOR ADDRESS Department of Physics, Loyola University Chicago, Chicago, IL 60660

ACS Paragon Plus Environment

1

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 36

ABSTRACT

Repetitive trinucleotide DNA sequences at specific genetic loci are associated with numerous hereditary, neurodegenerative diseases. The propensity of single-stranded domains containing these sequences to form secondary structure via extensive self-complementarity disrupts normal DNA processing to create genetic instabilities. To investigate these intrastrand structural dynamics, a DNA hairpin system was devised for single-molecule fluorescence study of the folding kinetics and energetics for secondary structure formation between two interacting, repetitive domains with specific numbers of the same trinucleotide motif (CXG), where X = T or A. Single-molecule FRET (smFRET) data show discrete conformational transitions between unstructured and closed hairpin states. The lifetimes of the closed hairpin states correlate with the number of repeats, with (CTG)N:(CTG)N domains maintaining longer-lived, closed states than equivalent-sized (CAG)N:(CAG)N domains. NaCl promotes similar degree of stabilization for the closed hairpin states of both repeat sequences. Temperature-based, smFRET experiments reveal that NaCl favors hairpin closing for (CAG)N:(CAG)N by pre-ordering single-stranded repeat domains to accelerate the closing transition. In contrast, NaCl slows the opening transition of CTG hairpins; however, it promotes misfolded conformations that require unfolding. Energy diagrams illustrate the distinct folding pathways of (CTG)N and (CAG)N repeat domains and identify features that may contribute to their gene-destabilizing effects.


2

Page 3 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


INTRODUCTION Numerous hereditary disorders arise from genomic instabilities associated with the expansion of small, repeating DNA motifs (microsatellites) at specific loci.1-4 Certain cancers also display increased frequency of expanded microsatellite domains.5-7 Individual microsatellites typically range in size from two to six nucleotides and may be located in coding and non-coding regions.8 Several microsatellite-related disorders involve the trinucleotide sequence (CXG)N (where X = A, C, G, or T), including Fragile X syndrome (CGG), various forms of spinocerebellar ataxia (CAG), and muscular dystrophy type 1 (CTG).3, 9, 10 The severity and onset age of these disorders correlate with the number of repeat units (N) comprising the sequence-specific, microsatellite domain.11 Expansion in the number of repeating units can occur spontaneously and intergenerationally through several mechanisms that involve DNA replication, transcription, or repair, indicating that microsatellite-related instabilities contribute to systemic disruption of DNArelated processes to induce pathogenic states.12-19 Inaccurate processing of repetitive, single-stranded DNA (e.g., misalignment due to polymerase slippage and flaps from repair synthesis) at these disease-linked loci can lead to the formation of non-helical, secondary structure. These structures have been identified as potential causative agents for pathogenic states that sequester unpaired repeat domains to prolong their lifetimes to physiologically relevant timescales and promote gene-destabilizing, expansion events.8, 18, 20 As to the structural nature of the repeat extrusions, small sizes exist primarily as unstructured loops,21-23 while larger domains form intrastrand hairpins, slip-outs, slipped strands, and higher order structures, such as triplexes and quadruplexes.9 Table 1 summarizes known repeat sequences and their reported in vitro secondary structures. Calorimetric and footprinting studies have systematically characterized the contributions of sequence and size to the stability and


3


Page 4 of 36

conformation of trinucleotide-repeat domains.24-28 The stability of these structures correlates with the number of interacting repeats.9, 26, 27, 29, 30 The propensity for single-stranded, CXG repeat domains to adopt stable secondary structure arises from extensive self-complementarity in which two C-G Watson-Crick base pairs form per pair of interacting triplets. The identity of the central mismatch contributes to the overall stability of the interacting triplets with G > T > A > C.9 The polymorphic capability of these repetitive sequences does not require extensive tracts as the “even/odd” rule shows that parity in the binding register can expand the number of possible conformations for relatively small hairpin sizes.31-33 Although the size threshold for the onset of diseases states associated with these repeat domains typically exceeds 30 repeats,11, 34 the in vivo condition of these domains and their fraction that participates in discrete expansion events remain unclear. While large expansion events that are on the order of the repeat tract can occur in postmitotic neurons35 and non-dividing cells,36 small extrusions may be targeted as expansion substrates. Recent in vivo studies reveal that these repeat tracts are dynamic, undergoing small expansion/contraction events (1 – 5 repeat units) with length-dependent frequency.37, 38 Expansionlinked, DNA repair mechanisms (transcription coupled repair, nucleotide excision repair, and base excision repair) process extrusions of similar size to these expansion events.11 In addition, tissue studies using three-way junction antibodies indicate that repeat domains may be populated with clusters of extrusions that range in size from 1 – 100 repeat units.39 In vitro functional studies implicate trinucleotide repeat hairpins with direct roles in the interference of multiple DNA repair pathways, such as nonproductive binding modes for the mismatch repair protein Msh2-Msh3,40, 41 inhibition of FEN1-mediated flap cleavage,42 and reduced endonucleolytic proficiency by APE1 in apurinic/apyrimidinic repair.43 Although significant advances have been made toward understanding expansion mechanisms and the resulting genetic instabilities, the underlying


4

Page 5 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


molecular interactions that influence these mechanisms remain unclear, including the behavior and conformation of secondary structure within the repeat tracts. The thermodynamic and structural studies of (CXG)N sequences with repeat numbers capable of extensive intrastrand contacts and hairpin formation indicate that they exist in a complex folding landscape and may adopt multiple conformations.24 In fact, studies on model DNA hairpins with fully complementary, single-register stems by single-molecule approaches,44-48 fluorescence correlation spectroscopy,49-51 and temperature jump methods52, 53 show that hairpins undergo rapid, continual opening and closing conformational transitions that are populated by metastable states and are susceptible to misfolding. A recent single-molecule study revealed hairpin polymorphism associated with the pentanucleotide repeat sequence TGGAA.33 For (CXG)N sequences, the effect of their repetitive nature, self-complementarity, and high mismatch content on the kinetics of their intrastrand folding behavior remains unknown. Here, the structural dynamics for hairpin formation via stem formation between two repetitive domains consisting of physiologically relevant sizes of CTG and CAG repeat sequences (6 – 12 repeats in total size) were investigated using singlemolecule fluorescence resonance energy transfer (smFRET). The closed hairpin states for both CAG and CTG sequences persist on the seconds scale with CTG hairpins longer lived than CAG hairpins. As expected, the lifetimes of CTG hairpins correlate with the number of repeats in equally sized (i.e., symmetric) domains involved in stem formation; however, asymmetry in the size of the interacting repeat domains alters the transition kinetics to reflect features of both domain sizes. In contrast to findings for model DNA hairpins,45 the opening and closing transition rates for equivalent-sized CTG and CAG hairpins vary in a distinct, sequence-dependent manner to changes in monovalent salt conditions. Equilibrium and transition-state analysis from temperature-based smFRET measurements reveal that folding of CTG and CAG hairpins becomes less exothermic


5


Page 6 of 36

with increasing monovalent salt; however, CAG hairpins exhibit pre-ordering to promote the closed hairpin state, while CTG hairpins folding encounter heightened enthalpic and entropic barriers that limit folding, most likely from misfolded species. The findings illustrate the dynamic behavior of structure formation by single-stranded, repetitive DNA domains and identify new aspects to these trinucleotide sequences that may contribute to the disruption of DNA processing mechanisms and lead to pathogenic outcomes.

EXPERIMENTAL SECTION Materials All oligonucleotides were purchased from Integrated DNA Technologies, Inc. (Coralville, IA) with PAGE purification. Each hairpin (HP) strand (CXG) N1, N2 consists of a 13-bp tethering sequence and two trinucleotide domains (CXG)N1 and (CXG)N2. The repeat domains are separated by an A15 linker, and the HP strand is 3’ labeled with the fluorescent dye Cy3. The following sequences were used: (CTG)3,3: 5’-CTCTGCGTCACCGTT(CTG)3A15(CTG)3/Cy3/-3’ (CTG)4,4: 5’-CTCTGCGTCACCGTT(CTG)4A15(CTG)4/Cy3/-3’ (CTG)5,5: 5’-CTCTGCGTCACCGTT(CTG)5A15(CTG)5/Cy3/-3’ (CTG)6,6: 5’-CTCTGCGTCACCGTT(CTG)6A15(CTG)6/Cy3/-3’ (CTG)5,4: 5’-CTCTGCGTCACCGTT(CTG)5A15(CTG)4/Cy3/-3’


6

Page 7 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(CAG)5,5: 5’-CTCTGCGTCACCGTT(CAG)5A15(CAG)5/Cy3/-3’ A tether strand (T) for immobilization is complementary to the 13-bp tethering domain of the HP strand. The T strand, internally labeled with Cy5 and 3’ biotinylated, has the sequence 5’CGGTGAC/Cy5/GCAGAGCACGT-3’/biotin. All constructs were prepared by annealing the appropriate strands at a HP:T ratio of 5:1 with a T strand concentration of 1 µM. The strands were heated to 94 °C for two minutes and then underwent slow, light-protected cooling to room temperature. The annealing conditions were 25 mM Tris-Cl (pH 8.0), 50 mM NaCl. The annealed DNA were kept on ice until measurement.

Single-molecule fluorescence microscopy The assembled DNA constructs were immobilized to quartz slides (Technical Glass, OH) functionalized through successive addition of biotinylated bovine serum albumin (0.5 mg/mL) and streptavidin (0.1 mg/mL). The DNA were immobilized at a tether concentration of 10 pM in the given NaCl concentration and 25 mM sodium cacodylate (pH 7.0). The experiments were performed by prism-based, total internal reflection fluorescence imaging with an Olympus IX-83 inverted microscope with a 60X water objective. The samples were separately excited with 532nm and 637-nm lasers (Coherent, Santa Clara, CA) to acquire the data and to confirm the presence of the Cy5 dye, respectively. To slow photobleaching, a deoxygenating imaging buffer comprising 25 mM sodium cacodylate (pH 7.0), 0.8 mM Trolox, 8% glucose, 0.1 mg/ml glucose oxidase, and 0.04 mg/ml catalase was used. Data were acquired at 10 Hz with an Andor iXon DU897 EMCCD (Andor Technologies, Belfast, UK) cooled to -60 °C. The emission signals from the Cy3 and Cy5 dyes were simultaneously collected for each field of molecules by spatial separation onto different regions of the EMCCD through a series of dichroic mirrors (640-nm cutoff; Chroma Technologies,


7


Page 8 of 36

Bellow Falls, VT). The emission signals corresponding to each molecule were identified using an affine transformation previously determined from a control slide coated with fluorescent nanobeads as fiducial markers. The emission intensities for each dye were corrected for local background. A FRET time trace was calculated for each molecule from the intensity time traces of the Cy3 and Cy5 signals as

𝐹𝑅𝐸𝑇(𝑡) =

𝐼𝐴(𝑡) ― 𝛽𝐼𝐷(𝑡) 𝐼𝐴(𝑡) + 𝐼𝐷(𝑡)

(1)

where β represents the correction factor for the fractional crossover of the Cy3 signal into the Cy5 channel. FRET histograms were generated from multiple fields of views and were best fit by Gaussian distributions using Levenberg-Marquardt optimization with OriginPro (OriginLab, Northampton, MA). Kinetic analysis of time traces was performed using custom software. Briefly, dwell-time histograms were constructed from the measured durations of individual open and closed hairpin states for hundreds of molecules in each condition. Cumulative histograms were generated so that the contribution of slower, more broadly distributed events could be determined. The open and closed histograms were best fit with single-exponential equations to determine the rate constants and lifetimes for the closing and opening transition states, respectively.54

Temperature-based smFRET experiments and analysis For temperature-based smFRET measurements, the assembled flow chambers were placed in a CHS-1 heated platform with a stage adaptor and connected to a single-channel TC-324C heater controller (Harvard Apparatus). The 60X objective was separately heated with an OWS-2 objective warmer (Harvard Apparatus). The chamber and objective temperatures were continually monitored with calibrated thermosistors. Data were acquired at temperatures ranging from 24 – 45


8

Page 9 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


°C with duplicate experiments performed for all conditions. The thermodynamic parameters (ΔHo and ΔSo) were determined from linear fits to

(𝑇1) ―∆𝑆𝑜

―𝑅𝑙𝑛[𝐾𝑒𝑞] = ∆𝐻𝑜

where 𝐾𝑒𝑞 was defined as 𝐾𝑒𝑞 =

𝑘𝑐

(2)

𝑘𝑜. The free energy change at 25 °C was calculated as Δ𝐺𝑜 = Δ𝐻𝑜 ―𝑇Δ𝑆𝑜

(3)

Transition-state analysis for the hairpins was performed by constructing Eyring plots from the opening and closing rates as a function of inverse temperature to determine the transition-state enthalpy (∆𝐻 ‡ ) and entropy (∆𝑆 ‡ ) values according to55, 56: ―𝑅𝑙𝑛[𝑘𝑜𝑣 ―1] = ∆𝐻𝑜‡

(𝑇1) ―∆𝑆𝑜‡

(4)

―𝑅𝑙𝑛[𝑘𝑐𝑣 ―1] = ∆𝐻𝑐‡

(𝑇1) ―∆𝑆𝑐‡

(5)

where ν represents the collision frequency, ν = 1013 s-1.57

RESULTS AND DISCUSSION Design of DNA hairpin constructs for smFRET experiments Hairpins are the primary, non-helical secondary structure formed by self-complementary, single-stranded DNA domains, such as the trinucleotide (CXG)N motifs. Its formation occurs primarily through a two-step process: 1) loop closure, which is the rate-limiting step that brings the ends of the loop together via conformational diffusive search,44,

49, 50, 52, 53

and 2) stem


9


Page 10 of 36

formation, which involves a rapid alignment mechanism (i.e., zippering) to achieve the correct binding register and maximize the number of base-pair interactions.58,

59

With appropriate

placement of Cy3 and Cy5 fluorescent dyes, continuous conformational transitions between the unstructured (open) and hairpin (closed) states can be probed by smFRET due to resulting changes in the interdye distance (Figure 1a). The work described herein addresses the kinetics of hairpin folding via stem formation between a pair of (CXG)N domains that vary in the number of interacting repeat units (3 – 6 repeats per domain) and in sequence (CAG and CTG). For these experiments, hairpin constructs were designed with a specific number of the same trinucleotide sequence divided into two repeat domains (Figure 1b). The nomenclature for the constructs is (CXG)N1,N2, where X represents the central nucleotide (A or T) and N1 and N2 refer to the number of repeat units in each domain. Because the lifetime of loop closure scales with loop size,50, 53 the two repeat domains are connected with a poly(dA) linker that increases the loop size from the physiological loop size of 3 – 4 nt to 15 nt. This linker size extends the loop closure time to permit real-time detection of the conformational transitions involving the trinucleotide domains. Surface immobilization of the hairpins for imaging permits direct measurement of (CXG)N1:(CXG)N2 interactions and minimizes interference from interstrand contacts.

Increasing repeat number promotes stem formation for symmetric, CTG-rich hairpins Symmetric CTG hairpins, which contain an equivalent number of CTG repeats in each domain (N1 = N2) that must interact for stem formation, were assembled, and the number of repeats in the domains were varied from 3 – 6 (Figures 1c - f). The smFRET time traces from individual molecules show stark differences in the transition behavior of the hairpins at 25 mM NaCl.


10

Page 11 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(CTG)3,3 remains almost exclusively in the low FRET state. The value of this low FRET state (0.24 ± 0.11) is consistent with the prediction from treating the single-stranded domain as an unstructured polymer with a low-salt persistence length of 3 nm.60 Together, these results indicate that (CTG)3,3 primarily exists in an unstructured conformation. As the number of repeats in the domains expands, the constructs increasingly populate high FRET states, which correspond to the closed hairpin via loop closure and stem formation. Time traces for (CTG)4,4 and (CTG)5,5 show rapid interconversion between the low and high FRET states, demonstrating that these sequences undergo continual conformational transitions between the unstructured and closed hairpin states. (CTG)6,6 produces a predominantly high FRET signal such that the unstructured state only occurs transiently, so it is primarily in the closed hairpin state. The smFRET histograms constructed from the accumulated FRET behavior of each symmetric domain construct show two distinct populations: a low FRET population that represents the open (unstructured) state and a high FRET population that corresponds to the closed hairpin (structured) state. The histograms confirm that the high FRET regime becomes increasingly populated as the number of interacting repeat domains in the stem expands, corroborating previous studies that CTG hairpin formation becomes more favorable with increasing repeat number.24, 25, 28 but also demonstrating the dynamic nature of these sequences. The opening and closing transition rates, based on dwell-time analysis, for all of the hairpins were best fit by single-exponential equations, consistent with observations from model hairpins (Figure S1a and b).45, 53, 61 Single-exponential fits to the high FRET dwell times show that the hairpin opening rate (kopening) slows with increasing repeat number from 2.98 s-1 for (CTG)4,4 to 2.25 x 10-2 s-1 for (CTG)6,6. This change represents a ~130-fold increase in the lifetime of the closed state from 0.34 s to 44.4 s, reflecting the stabilizing effect of two additional interacting sets


11


Page 12 of 36

of triplets in the stem. In contrast, the lifetime of the open state contracts with increasing number of repeat domains, but the acceleration of the closing transition with increasing number of repeats is nearly 30-fold smaller than the slowing of the opening transition, with kclosing increasing from 0.37 s-1 for (CTG)4,4 to 1.83 s-1 for (CTG)6,6. kclosing scales exponentially with the total number of nucleotides per repeat domain, suggesting that the energetic barrier for the closing transition decreases as the number of repeats expands. Single-molecule investigations of hairpin closure have reported diverging results on the influence of stem sequence on closing transitions.45, 62 Here, the observed closing rates are significantly slower than expected for poly(dA) loops of this size, based on comparison with larger loop sizes.45, 50, 61 This finding agrees with earlier work that showed slower duplex formation by (CTG)25 strands compared to model duplexes, suggesting that repetitive content impedes secondary structure formation.12 The transitional barrier associated with stem formation due to the repetitive sequence and high mismatch content appears to outweigh the conformational diffusion barrier for loop closure such that stem formation, not loop closure, acts as the rate-limiting step for the hairpin closing transition. Interactions between these repeat domains likely have a lower probability of productive encounters (i.e., no stem formation) than single-register stems capable of full base pairing. Larger repeat number may enhance the probability of productive interdomain encounters to promote zippering of the stem, perhaps through formation of a greater number of initial contacts.59 Calculation of equilibrium constants between the closed and open state (Keq) according to 𝐾𝑒𝑞 =

𝑘𝑐𝑙𝑜𝑠𝑖𝑛𝑔

-3 𝑘𝑜𝑝𝑒𝑛𝑖𝑛𝑔 shows that Keq increases from 4.5 x 10 for (CTG)3,3 to 81 for (CTG)6,6,

an 18,000-fold increase. Each additional repeat unit increase Keq by ~ 26-fold. Free energies (ΔGo) for each construct calculated from the Keq values (Figure S1) increase linearly with the number of repeats in each domain with a slope of 1.94 kcal/mol/repeat, which agrees with previous


12

Page 13 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


calorimetric findings25 and with nearest neighbor predictions for the free energy difference per repeat,63 demonstrating that the extended loop does not disrupt the length-dependent energetics of the symmetric (CTG)N:(CTG)N interactions in hairpin formation. The increasing stability of the hairpins with growing repeat number reflects two favorable processes for the closed hairpin state: slowing of the opening transition with the presence of additional triplets and acceleration of the closing transition through higher likelihood of a productive encounter between the repeat domains.

Asymmetric repeat domains accelerate hairpin formation Hairpin stability increases with repeat number in an irregular pattern due to parity in the binding register as a result of asymmetry in the stem, which can induce multiple hairpin conformations.31-33 Single-molecule investigation of the pentanucleotide (CTGCT)N showed distinct FRET states corresponding due to different binding registers.33 To examine the role of dissimilar repeat domains on hairpin formation, an asymmetric construct was designed with two differing domain sizes, (CTG)5,4, and compared with the symmetric constructs, (CTG)4,4 and (CTG)5,5. The experiments were conducted at higher salt conditions to favor the closed hairpin state. At near physiological NaCl concentration (100 mM), the symmetric constructs (CTG)4,4 and (CTG)5,5 exhibit clearly resolvable opening and closing transitions, and smFRET histograms show bimodal distributions in which the closed states for both constructs become increasingly populated relative to the observations at 25 mM NaCl (Figures 2a and b). Kinetic analysis reveals that the shift toward the closed hairpin conformation occurs through kopening slowing approximately fivefold for both constructs and not from an acceleration in the closing rate (Figure S2). (CTG)5,5 populates the unstructured state for briefer times than (CTG)4,4, consistent with the low-salt


13


Page 14 of 36

observations. The relative insensitivity of kclosing to NaCl is addressed further in later sections. As a result, Keq increases from 0.1 to 0.5 for (CTG)4,4 and from 3.5 to 22.6 for (CTG)5,5, with NaCl changing from 25 to 100 mM. So, the closed states become more energetically favorable with increasing salt, and the additional NaCl provides similar net stabilization of ~ 1 kcal/mol for both constructs. For the asymmetric construct (CTG)5,4, its smFRET histogram, in similar fashion to the symmetric (CTG)4,4 and (CTG)5,5 constructs, shows a bimodal population, and it has a high FRET state population that is larger than that of (CTG)4,4 but smaller than (CTG)5,5. In contrast to the symmetric constructs in which energetic differences arose from longer-lived closed states, (CTG)4,4 and (CTG)5,4 have similar kopening values that reflect both constructs containing stems with a maximum of four interacting repeat units. The additional repeat unit accelerates kclosing from 0.33 s-1 for (CTG)4,4 to 1.02 s-1. The basis for the faster closing rate may be that the (CTG)5 domain presents two different binding registers that are energetically equivalent for (CTG)4, which increases the probability of a productive encounter and stem formation. Overall, (CTG)5,4 has a Keq of ~ 1.5, which is 3-fold greater than (CTG)4,4 and 15-fold smaller than (CTG)5,5, consistent with previously observed irregularities due to differences of a single repeat unit.25, 32 The (CTG)5 domain of the (CTG)5,4 construct presents two equivalently sized, potential binding registers for the (CTG)4 domain, denoted as (0) and (-3) for their distance from the 5’ end of the (CTG)5 domain. The two registers suggests that multiple high FRET states may be populated, representing different hairpin conformations with similar stabilities from four sets of interacting triplets. The high FRET populations for the symmetric constructs are best fit by singlepeak Gaussians with similar values (median ± width) of 0.78 ± 0.22 and 0.80 ± 0.20 for (CTG)4,4 and (CTG)5,5, respectively. Similar analysis for (CTG)5,4 gives a lower value of 0.71 ± 0.24,


14

Page 15 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


analogous to the lower FRET shift observed for (CTGCT)N.33 This lower FRET value may arise from two scenarios: i) rapid transitions between the (0) and (-3) registers that appear as a single population or ii) two stably populated binding registers. To discern between these two scenarios, histograms were constructed from the mean values of each observed, closed state for (CTG)4,4, (CTG)5,4, and (CTG)5,5 to determine whether the width of the Gaussian distributions from the FRET histograms arise from fluctuations within the closed state (narrower distribution of mean FRET states) or differences in the binding register (broader distribution) (Figure S3). The two symmetric constructs display narrower distributions, which is consistent with a single register conformation. (CTG)5,4 has a slightly broader distribution that can be resolved into two populations with widths similar to the single-register symmetric cases (FRET ~ 0.65 and 0.74) that may represent the (-3) and (0) binding registers, respectively, with the (-3) register having a lower mean FRET signal because of the 3-nt increase in distance. Overall, the CTG constructs give no clear indication of alternate closed conformations arising from different registers in terms of discrete FRET transitions or multiphasic kinetics. These findings suggest that once sufficient contacts have been established to nucleate the stem, any partially aligned or misfolded intermediates do not persist on our observation timescale and are rapidly resolved to the fully-paired conformation by some rearrangement mechanism, such as internal displacement59 or a slithering.64

NaCl concentration alters the transition kinetics for the (CTG)5,5 hairpin The preceding experiments demonstrate that changes in NaCl concentration may affect (CTG)N hairpin transition rates differently than hairpins with fully paired, non-repetitive stems.45 The (CTG)5,5 construct was selected for further analysis of salt effects on hairpin behavior, and its


15


Page 16 of 36

conformational transitions were measured in different NaCl conditions, ranging from 0 mM (buffer only) to 1 M. A smFRET heat map was constructed from the individual histograms for each NaCl concentration (Figure 3a). As the NaCl concentration increases, the FRET population shifts from predominantly low FRET to high FRET, indicating that the increasing number of monovalent cations promotes the closed conformation, consistent with expected stabilization of secondary structure by monovalent cations. As a first step to explore the origins of this effect, the opening and closing rates were measured at each salt concentration (Figure 3b). The opening rate slows nearly 500-fold with increasing salt concentration, decreasing from 0.80 s-1 to 2.0 x 10-3 s-1. This slowing indicates that salt stabilizes (CTG)5:(CTG)5 interactions within the stem. The closing rate increases modestly from 0.36 s-1 in buffer only conditions and plateaus at approximately 1.1 s-1 for NaCl concentrations greater than ~100 mM. This finding deviates from model hairpins in that NaCl weakly influences the closing transition, which may be a consequence of NaCl-stabilization of misfolded structures that interfere with the closing transition. Keq increases from approximately 0.5 in 0 mM NaCl to 650 in 1 M NaCl, a ~1300-fold change toward the closed hairpin state, in agreement with the trend observed in the FRET heat map. The plot of ln(Keq) versus ln([Na+]) shows a strong linear characteristic (Figure 3c), and the slope (ΔΓ = 1.88) provides a measure of the net number of uptaken and released cations during formation and dissolution of secondary structure.55, 65, 66

Identity of central mismatch switches the dominant rate for hairpin formation The central mismatch in CXG repeats has a pronounced effect on the structural energetics as (CAG)N hairpins are less stable than (CTG)N hairpins with the difference arising from the


16

Page 17 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


stabilizing effect of two hydrogen bonds present in two possible T·T base-pairing modes within the CTG/GTC context, compared to a single hydrogen bond present in A·A mismatches.67 Here, to compare the effect of the central mismatch on CXG hairpin folding behavior, a (CAG)5,5 construct was used with A replacing T in the central position of the trinucleotide repeats. The smFRET heat map shows that (CAG)5,5 undergoes a similar FRET shift as (CTG)5,5 to favor the closed conformation with increasing NaCl (Figure 4a). At low monovalent concentrations, (CAG)5,5 exists in a predominantly low FRET state, indicative of the unstructured conformation. As the salt concentration increases, the high FRET state becomes increasingly populated, although the low FRET state remains more populated for (CAG)5,5 than for (CTG)5,5. For (CAG)5,5, kopening slows 10-fold from 6.0 s-1 in buffer-only conditions to 0.58 s-1 in 1 M NaCl, while kclosing increases from 0.038 to 2.78 s-1 over the same NaCl range. The degree to which kopening decreases is 50-fold smaller than the kopening for (CTG)5,5. In contrast, kclosing for (CAG)5,5 undergoes a 70-fold acceleration, compared to 3-fold for (CTG)5,5 due to the change in NaCl conditions. Equilibrium constant calculations show that the Keq for (CAG)5,5 increases from 0.006 to 4.9, a near 800-fold shift, similar to that observed for (CTG)5. The Keq values for (CAG)5,5 are generally 100-fold lower than the Keq values for (CTG)5,5 at all NaCl concentrations. The plot of ln(Keq) as a function of ln([Na+]) for (CAG)5,5 also has a strong linear characteristic with a ΔΓ value of 1.94, indicating that despite differences in sequence and kinetic behavior, the structural transitions for these samesized constructs involve a similar change in the number of non-site-specific cations for folding. To examine the role of loop composition on the folding of the (CTG)5,5 and (CAG)5,5 constructs, experiments were conducted with a poly(dT) linker (T15) replacing the poly(dA) linker (Table S1). The similarity of the opening rates for the two linkers indicates that it does not measurably influence the stability of the closed state. The constructs with the poly(dT) linker exhibit faster


17


Page 18 of 36

closing rates, which reflect its smaller transition barrier due to a lower stacking potential compared to the poly(dA) linker.50, 52, 68, 69 Both linkers show similar sequence-dependent changes in the rates with NaCl, indicating that the relative differences in the behavior of the hairpins correspond to features of the folding transitions involving the repeat domains.

Temperature-dependent analysis indicates salt-dependent, alternate folding of (CTG)5,5 Although (CTG)5,5 and (CAG)5,5 undergo similar shifts in conformational equilibria with increasing NaCl, the differences in the evolution of the transition rates suggest that these sequences encounter distinct barriers to folding. Temperature-based smFRET experiments were performed to dissect the entropic and enthalpic components of the folding pathways for these sequences. The open and closing rates for (CTG)5,5 and (CAG)5,5 were measured over a range of temperatures at three different NaCl concentrations (25, 100 and 500 mM). The rate-derived Keq values for (CTG)5,5 and (CAG)5,5 were then plotted as functions of the inverse temperature for the three NaCl concentrations (Figure 5a and Figure S4a, respectively). From the linear plots, Table 2 shows the equilibrium thermodynamic parameters (ΔGo, ΔHo, and ΔSo) for (CTG)5,5 and (CAG)5,5 that were extracted by van’t Hoff analysis (Eq. 2) at the each NaCl condition. Consistent with the FRET heat maps, (CTG)5,5 and (CAG)5,5 show similar changes in free energy between 25 and 500 mM NaCl to favor the closed conformation with ΔΔGo values of -2.25 and -3.07 kcal/mol, respectively. Although both sequences have negative slopes in the van’t Hoff analysis that correspond to favorable energy release in hairpin folding due to base pairing in stem formation, CTG hairpin formation becomes significantly less exothermic with increasing NaCl (ΔΔHo = +15.85 kcal/mol), while a more modest ΔΔHo = +4.43 kcal/mol is seen for CAG. A larger reduction in the entropic


18

Page 19 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


cost with ΔΔSo = +60.73 cal/mol/K offsets the reduced exothermicity for (CTG)5,5 to promote hairpin formation in increasing NaCl. (CAG)5,5 also shows increased ordering with ΔΔSo = +25.16 cal/mol/K. Overall, these findings are consistent with observations from duplex formation with full W-C base pairing in that significant pre-ordering occurs with increased NaCl to promote base pairing and compensate for decreased exothermicity.55 Strikingly, for (CTG)5,5, the enthalpic and entropic differences between the open and closed hairpin states approach zero as NaCl increases, additional evidence that the open conformation is becoming increasingly ordered with stabilizing interactions. Transition-state analysis was performed to quantify the NaCl-dependent energetic barriers involved in the transitions between the unstructured and closed hairpin states. Eyring plots constructed from the temperature-dependent rate data are shown in Figures 5b,c and Figures S4b,c for (CTG)5,5 and (CAG)5,5, respectively. Table 3 summarizes the NaCl-dependent, opening and closing transition-state thermodynamic parameters (ΔH ǂ and ΔS ǂ) for both sequences by applying Eqs. 4 and 5 to the linear profiles. Using the values in Tables 2 and 3, the energy diagrams in Figure 6 summarize the salt-dependent modulation of folding by the CTG and CAG hairpins. (CTG)5,5 encounters a growing enthalpic barrier for the closing transition that increases by 7.7 kcal/mol between 25 and 500 mM NaCl, compared to 2.2 kcal/mol for (CAG)5,5. Unstacking of the poly(dA) loop contributes to this enthalpic barrier for loop closure for both sequences,50, 52, 68 but the pronounced difference between the two sequences indicates the involvement of additional factors arising from the repeat domains. These factors likely included misfolded states, such as intradomain (CTG)2:(CTG)2 interactions with a CTG loop, that impede hairpin formation for (CTG)5,5. The presence of such extensive, alternative interactions would need to be broken for complete stem formation to occur between the two (CTG)5 domains. These alternative contacts


19


Page 20 of 36

would also expectedly reduce entropy. Indeed, Figure 6c shows that a more disordered transition state for (CTG)5,5 must occur in order to fold to the closed hairpin conformation with the full (CTG)5:(CTG)5 stem. In contrast, (CAG)5,5 appears less susceptible to misfolding because of the destabilizing A·A central mismatch. The lower entropy for the unstructured state with increased NaCl suggests that the single-stranded CAG domains become increasingly ordered or preorganized prior to stem formation, which reduces the entropic cost (Figure 6f). This preorganization step may increase the number of initial contacts to facilitate zippering for stem formation and accelerate hairpin formation. For the size of interacting trinucleotide domains reported here, a two-state model describes the observed folding process between the open and closed hairpin states. For the CTG domain sizes in this study, a greater number of repeats accelerates the closing rate, but increasing NaCl only weakly accelerates its closing transition, which may reflect stabilized misfolded states. The folding of longer repeat sizes likely includes longer-lived intermediates and misaligned registers that have sufficient stability to persist for observable timescales. There may be a sequence-dependent, size limit such that increasingly stable misfolded states arising from a larger number of interacting repeats could lead to slower closing rates and an increasing level of structural heterogeneity. This structural heterogeneity could interfere with the DNA repair proteins that target these domains. The findings from this study will contribute to the understanding of the structural dynamics for larger repetitive sequences as the interactions and folding pathways described herein may serve as building blocks for higher order structures.

CONCLUSIONS


20

Page 21 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


The work presented here demonstrates that single-stranded domains containing the trinucleotide repeat motif CXG, associated with numerous hereditary neurodegenerative disorders, display sequence-dependent kinetics in their folding behavior. Intrastrand CXG:CXG interactions involving small number of repeats can extend the lifetimes of secondary structure formation to physiologically relevant timescales. The susceptibility of these sequences, particularly CTG, to metastable states and misfolding due to their repetitive content, self-complementarity, and high mismatch content complicates the folding transitions. Because these repetitive domains serve as substrates in expansion mechanisms, this complex folding behavior may interfere with the targeting and processing of these sequences by kinetically regulating protein·DNA interactions. A fuller understanding of the structural dynamics and folding pathways of these sequences may also assist in the development of rationally designed small molecules for targeting extruded domains with these sequences in diagnostic and therapeutic applications.70, 71

AUTHOR INFORMATION Corresponding Author *E-mail: [email protected]. Phone: (773) 508-8478. Fax: (773) 508-3534 Author Contributions Experiments performed by M.L.M., M.P.L., T.T., and R.S.S. Manuscript prepared by M.L.M. and B.C.

ABBREVIATIONS FRET, fluorescence resonance energy transfer; nt, nucleotide; bp, base pair; WC, Watson-Crick


21


Page 22 of 36

ACKNOWLEDGEMENTS This work was supported by Mulcahy Scholarship to M.L.M. and start-up funds from Loyola University Chicago to B.C.

SUPPORTING INFORMATION AVAILABLE Figure S1: Plots of free energy calculations and transition kinetics as functions of the number of symmetric repeats. Figure S2: Dwell-time histograms showing the influence of asymmetry in interacting repeat domain size on hairpin transition kinetics and conformation. Figure S3: Closedstate histogram analysis on effect of domain asymmetry. Figure S4: Transition-state analysis of temperature-based smFRET measurements of the opening and closing kinetics for (CAG)5,5 in varying NaCl. Table S1: Summary of results for hairpin constructs with poly d(T) linker. This information is available free of charge via the Internet at http://pubs.acs.org.

REFERENCES 1. Hagerman, K. A.; Ruan, H.; Edamura, K. N.; Matsuura, T.; Pearson, C. E.; Wang, Y. H. The ATTCT Repeats of Spinocerebellar Ataxia Type 10 Display Strong Nucleosome Assembly Which is Enhanced by Repeat Interruptions. Gene 2009, 434, 29-34. 2. Haeusler, A. R.; Donnelly, C. J.; Periz, G.; Simko, E. A.; Shaw, P. G.; Kim, M. S.; Maragakis, N. J.; Troncoso, J. C.; Pandey, A.; Sattler, R.; et al. C9orf72 Nucleotide Repeat Structures Initiate Molecular Cascades of Disease. Nature 2014, 507, 195-200. 3. Hagerman, P. J.; Hagerman, R. J. Fragile X-Associated Tremor/Ataxia Syndrome. Ann. N. Y. Acad. Sci. 2015, 1338, 58-70. 4. Jayadev, S.; Bird, T. D. Hereditary Ataxias: Overview. Genet. Med. 2013, 15, 673-83. 5. Kim, T. M.; Laird, P. W.; Park, P. J. The Landscape of Microsatellite Instability in Colorectal and Endometrial Cancer Genomes. Cell 2013, 155, 858-68.


22

Page 23 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


6. Watson, M. M.; Berg, M.; Søreide, K. Prevalence and Implications of Elevated Microsatellite Alterations at Selected Tetranucleotides in Cancer. Br. J. Cancer 2014, 111, 823-7. 7. Hause, R. J.; Pritchard, C. C.; Shendure, J.; Salipante, S. J. Classification and Characterization of Microsatellite Instability Across 18 Cancer Types. Nat. Med. 2016, 22, 1342135. 8. McMurray, C. T. Mechanisms of Trinucleotide Repeat Instability During Human Development. Nat. Rev. Genet. 2010, 11, 786-99. 9. Mirkin, S. M. DNA Structures, Repeat Expansions and Human Hereditary Disorders. Curr. Opin. Struct. Biol. 2006, 16, 351-8. 10. Dueñas, A. M.; Goold, R.; Giunti, P. Molecular Pathogenesis of Spinocerebellar Ataxias. Brain 2006, 129, 1357-70. 11. Lee, D. Y.; McMurray, C. T. Trinucleotide Expansion in Disease: Why Is There a Length Threshold? Curr. Opin. Genet. Dev. 2014, 26, 131-40. 12. Gacy, A. M.; McMurray, C. T. Influence of Hairpins on Template Reannealing at Trinucleotide Repeat Duplexes: A Model for Slipped DNA. Biochemistry 1998, 37, 9426-34. 13. Hartenstine, M. J.; Goodman, M. F.; Petruska, J. Weak Strand Displacement Activity enables Human DNA Polymerase Beta to Expand CAG/CTG Triplet Repeats at Strand Breaks. J Biol. Chem. 2002, 277, 41379-89. 14. Kovtun, I. V.; McMurray, C. T. Features of Trinucleotide Repeat Instability In Vivo. Cell Res. 2008, 18, 198-213. 15. Jarem, D. A.; Wilson, N. R.; Delaney, S. Structure-Dependent DNA Damage and Repair in a Trinucleotide Repeat Sequence. Biochemistry 2009, 48, 6655-63. 16. Crespan, E.; Czabany, T.; Maga, G.; Hübscher, U. Microhomology-Mediated DNA Strand Annealing and Elongation by Human DNA Polymerases λ and β on Normal and Repetitive DNA Sequences. Nucleic Acids Res. 2012, 40, 5577-90. 17. Liu, Y.; Wilson, S. H. DNA Base Excision Repair: A Mechanism of Trinucleotide Repeat Expansion. Trends Biochem. Sci. 2012, 37, 162-72. 18. Zhao, X. N.; Usdin, K. The Repeat Expansion Diseases: The Dark Side of DNA Repair. DNA Repair (Amst.) 2015, 32, 96-105. 19. Zhao, X. N.; Lokanga, R.; Allette, K.; Gazy, I.; Wu, D.; Usdin, K. A MutSβ-Dependent Contribution of MutSα to Repeat Expansions in Fragile X Premutation Mice? PLoS Genet. 2016, 12, e1006190. 20. Mirkin, S. M. Expandable DNA Repeats and Human Disease. Nature 2007, 447, 932-40. 21. Völker, J.; Makube, N.; Plum, G. E.; Klump, H. H.; Breslauer, K. J. Conformational Energetics of Stable and Metastable States Formed by DNA Triplet Repeat Oligonucleotides: Implications for Triplet Expansion Diseases. Proc. Natl. Acad. Sci. U. S. A. 2002, 99, 14700-5. 22. Lee, B. J.; Barch, M.; Castner, E. W.; Völker, J.; Breslauer, K. J. Structure and Dynamics in DNA Looped Domains: CAG Triplet Repeat Sequence Dynamics Probed by 2-Aminopurine Fluorescence. Biochemistry 2007, 46, 10756-66. 23. Völker, J.; Plum, G. E.; Gindikin, V.; Klump, H. H.; Breslauer, K. J. Impact of Bulge Loop Size on DNA Triplet Repeat Domains: Implications for DNA Repair and Expansion. Biopolymers 2014, 101, 1-12. 24. Amrane, S.; Saccà, B.; Mills, M.; Chauhan, M.; Klump, H. H.; Mergny, J. L. LengthDependent Energetics of (CTG)n and (CAG)n Trinucleotide Repeats. Nucleic Acids Res. 2005, 33, 4065-77.


23


Page 24 of 36

25. Huang, J.; Delaney, S. Unique Length-Dependent Biophysical Properties of Repetitive DNA. J. Phys. Chem. B 2016, 120, 4195-203. 26. Pearson, C. E.; Sinden, R. R. Trinucleotide Repeat DNA Structures: Dynamic Mutations from Dynamic DNA. Curr. Opin. Struct. Biol. 1998, 8, 321-30. 27. Gacy, A. M.; Goellner, G.; Juranić, N.; Macura, S.; McMurray, C. T. Trinucleotide Repeats that Expand in Human Disease Form Hairpin Structures In Vitro. Cell 1995, 81, 533-40. 28. Paiva, A. M.; Sheardy, R. D. Influence of Sequence Context and Length on the Structure and Stability of Triplet Repeat DNA Oligomers. Biochemistry 2004, 43, 14218-27. 29. Lee, I. B.; Hong, S. C.; Lee, N. K.; Johner, A. Kinetics of the Triplex-Duplex Transition in DNA. Biophys. J. 2012, 103, 2492-501. 30. Usdin, K. NGG-Triplet Repeats Form Similar Intrastrand Structures: Implications for the Triplet Expansion Diseases. Nucleic Acids Res. 1998, 26, 4078-85. 31. Figueroa, A. A.; Cattie, D.; Delaney, S. Structure of Even/Odd Trinucleotide Repeat Sequences Modulates Persistence of Non-B Conformations and Conversion to Duplex. Biochemistry 2011, 50, 4441-50. 32. Chi, L. M.; Lam, S. L. Structural Roles of CTG Repeats in Slippage Expansion During DNA Replication. Nucleic Acids Res. 2005, 33, 1604-17. 33. Huang, T. Y.; Chang, C. K.; Kao, Y. F.; Chin, C. H.; Ni, C. W.; Hsu, H. Y.; Hu, N. J.; Hsieh, L. C.; Chou, S. H.; Lee, I. R.; et al. Parity-Dependent Hairpin Configurations of Repetitive DNA Sequence Promote Slippage Associated with DNA Expansion. Proc. Natl. Acad. Sci. U. S. A. 2017, 114, 9535-9540. 34. Usdin, K.; House, N. C.; Freudenreich, C. H. Repeat Instability During DNA Repair: Insights from Model Systems. Crit. Rev. Biochem. Mol. Biol. 2015, 50, 142-67. 35. Gonitel, R.; Moffitt, H.; Sathasivam, K.; Woodman, B.; Detloff, P. J.; Faull, R. L.; Bates, G. P. DNA Instability in Postmitotic Neurons. Proc. Natl. Acad. Sci. U. S. A. 2008, 105, 3467-72. 36. De Temmerman, N.; Sermon, K.; Seneca, S.; De Rycke, M.; Hilven, P.; Lissens, W.; Van Steirteghem, A.; Liebaers, I. Intergenerational Instability of the Expanded CTG Repeat in the DMPK Gene: Studies in Human Gametes and Preimplantation Embryos. Am. J. Hum. Genet. 2004, 75, 325-9. 37. Williams, G. M.; Surtees, J. A. MSH3 Promotes Dynamic Behavior of Trinucleotide Repeat Tracts In Vivo. Genetics 2015, 200, 737-54. 38. Kantartzis, A.; Williams, G. M.; Balakrishnan, L.; Roberts, R. L.; Surtees, J. A.; Bambara, R. A. Msh2-Msh3 Interferes with Okazaki Fragment Processing to Promote Trinucleotide Repeat Expansions. Cell Rep. 2012, 2, 216-22. 39. Axford, M. M.; Wang, Y. H.; Nakamori, M.; Zannis-Hadjopoulos, M.; Thornton, C. A.; Pearson, C. E. Detection of Slipped-DNAs at the Trinucleotide Repeats of the Myotonic Dystrophy Type I Disease Locus in Patient Tissues. PLoS Genet. 2013, 9, e1003866. 40. Lang, W. H.; Coats, J. E.; Majka, J.; Hura, G. L.; Lin, Y.; Rasnik, I.; McMurray, C. T. Conformational Trapping of Mismatch Recognition Complex MSH2/MSH3 on Repair-Resistant DNA Loops. Proc. Natl. Acad. Sci. U. S. A. 2011, 108, E837-44. 41. Owen, B. A.; Yang, Z.; Lai, M.; Gajec, M.; Gajek, M.; Badger, J. D.; Hayes, J. J.; Edelmann, W.; Kucherlapati, R.; Wilson, T. M.; et al. (CAG)(n)-Hairpin DNA Binds to Msh2Msh3 and Changes Properties of Mismatch Recognition. Nat. Struct. Mol. Biol. 2005, 12, 663-70. 42. Spiro, C.; Pelletier, R.; Rolfsmeier, M. L.; Dixon, M. J.; Lahue, R. S.; Gupta, G.; Park, M. S.; Chen, X.; Mariappan, S. V.; McMurray, C. T. Inhibition of FEN-1 Processing by DNA Secondary Structure at Trinucleotide Repeats. Mol. Cell 1999, 4, 1079-85.


24

Page 25 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


43. Li, M.; Völker, J.; Breslauer, K. J.; Wilson, D. M. APE1 Incision Activity at Abasic Sites in Tandem Repeat Sequences. J. Mol. Biol. 2014, 426, 2183-98. 44. Woodside, M. T.; Anthony, P. C.; Behnke-Parks, W. M.; Larizadeh, K.; Herschlag, D.; Block, S. M. Direct Measurement of the Full, Sequence-Dependent Folding Landscape of a Nucleic Acid. Science 2006, 314, 1001-4. 45. Tsukanov, R.; Tomov, T. E.; Berger, Y.; Liber, M.; Nir, E. Conformational Dynamics of DNA Hairpins at Millisecond Resolution Obtained from Analysis of Single-Molecule FRET Histograms. J. Phys. Chem. B 2013, 117, 16105-9. 46. Tsukanov, R.; Tomov, T. E.; Masoud, R.; Drory, H.; Plavner, N.; Liber, M.; Nir, E. Detailed study of DNA Hairpin Dynamics Using Single-Molecule Fluorescence Assisted by DNA Origami. J. Phys. Chem. B 2013, 117, 11932-42. 47. Deniz, A. A.; Dahan, M.; Grunwell, J. R.; Ha, T.; Faulhaber, A. E.; Chemla, D. S.; Weiss, S.; Schultz, P. G. Single-Pair Fluorescence Resonance Energy Transfer on Freely Diffusing Molecules: Observation of Förster Distance Dependence and Subpopulations. Proc. Natl. Acad. Sci. U. S. A. 1999, 96, 3670-5. 48. Grunwell, J. R.; Glass, J. L.; Lacoste, T. D.; Deniz, A. A.; Chemla, D. S.; Schultz, P. G. Monitoring the Conformational Fluctuations of DNA Hairpins Using Single-Pair Fluorescence Resonance Energy Transfer. J. Am. Chem. Soc. 2001, 123, 4295-303. 49. Wallace, M. I.; Ying, L.; Balasubramanian, S.; Klenerman, D. Non-Arrhenius Kinetics for the Loop Closure of a DNA Hairpin. Proc. Natl. Acad. Sci. U. S. A. 2001, 98, 5584-9. 50. Goddard, N. L.; Bonnet, G.; Krichevsky, O.; Libchaber, A. Sequence Dependent Rigidity of Single Stranded DNA. Phys. Rev. Lett. 2000, 85, 2400-3. 51. Mosayebi, M.; Romano, F.; Ouldridge, T. E.; Louis, A. A.; Doye, J. P. The Role of Loop Stacking in the Dynamics of DNA Hairpin Formation. J. Phys. Chem. B 2014, 118, 14326-35. 52. Ansari, A.; Shen, Y.; Kuznetsov, S. V. Misfolded Loops Decrease the Effective Rate of DNA Hairpin Formation. Phys. Rev. Lett. 2002, 88, 069801. 53. Kuznetsov, S. V.; Ren, C. C.; Woodson, S. A.; Ansari, A. Loop Dependence of the Stability and Dynamics of Nucleic Acid Hairpins. Nucleic Acids Res. 2008, 36, 1098-112. 54. Roy, R.; Hohng, S.; Ha, T. A Practical Guide to Single-Molecule FRET. Nat. Methods 2008, 5, 507-16. 55. Dupuis, N. F.; Holmstrom, E. D.; Nesbitt, D. J. Single-Molecule Kinetics Reveal CationPromoted DNA Duplex Formation through Ordering of Single-Stranded Helices. Biophys. J. 2013, 105, 756-66. 56. Holmstrom, E. D.; Nesbitt, D. J. Biophysical Insights from Temperature-Dependent Single-Molecule Förster Resonance Energy Transfer. Annu. Rev. Phys. Chem. 2016, 67, 441-65. 57. Holmstrom, E. D.; Nesbitt, D. J. Single-Molecule Fluorescence Resonance Energy Transfer Studies of the Human Telomerase RNA Pseudoknot: Temperature-/Urea-Dependent Folding Kinetics and Thermodynamics. J. Phys. Chem. B 2014, 118, 3853-63. 58. Kuznetsov, S. V.; Ansari, A. A Kinetic Zipper Model with Intrachain Interactions Applied to Nucleic Acid Hairpin Folding Kinetics. Biophys. J. 2012, 102, 101-11. 59. Ouldridge, T. E.; Sulc, P.; Romano, F.; Doye, J. P.; Louis, A. A. DNA Hybridization Kinetics: Zippering, Internal Displacement and Sequence Dependence. Nucleic Acids Res. 2013, 41, 8886-95. 60. Murphy, M. C.; Rasnik, I.; Cheng, W.; Lohman, T. M.; Ha, T. Probing Single-Stranded DNA Conformational Flexibility Using Fluorescence Spectroscopy. Biophys. J. 2004, 86, 2530-7.


25


Page 26 of 36

61. Baltierra-Jasso, L. E.; Morten, M. J.; Laflör, L.; Quinn, S. D.; Magennis, S. W. Crowdinginduced Hybridization of Single DNA Hairpins. J. Am. Chem. Soc. 2015, 137, 16020-3. 62. Neupane, K.; Wang, F.; Woodside, M. T. Direct Measurement of Sequence-Dependent Transition Path Times and Conformational Diffusion in DNA Duplex Formation. Proc. Natl. Acad. Sci. U. S. A. 2017, 114, 1329-1334. 63. Markham, N. R.; Zuker, M. UNAFold: Software for Nucleic Acid Folding and Hybridization. Methods Mol. Biol. 2008, 453, 3-31. 64. Sambriski, E. J.; Schwartz, D. C.; de Pablo, J. J. Uncovering Pathways in DNA Oligonucleotide Hybridization via Transition State Analysis. Proc. Natl. Acad. Sci. U. S. A. 2009, 106, 18125-30. 65. Bokinsky, G.; Rueda, D.; Misra, V. K.; Rhodes, M. M.; Gordus, A.; Babcock, H. P.; Walter, N. G.; Zhuang, X. Single-Molecule Transition-State Analysis of RNA Folding. Proc. Natl. Acad. Sci. U. S. A. 2003, 100, 9302-7. 66. Leipply, D.; Lambert, D.; Draper, D. E. Ion-RNA Interactions Thermodynamic Analysis of the Effects of Mono- and Divalent Ions on RNA Conformational Equilibria. Methods Enzymol. 2009, 469, 433-63. 67. He, G.; Kwok, C. K.; Lam, S. L. Preferential Base Pairing Modes of T·T Mismatches. FEBS Lett. 2011, 585, 3953-8. 68. Chen, W. S.; Chen, W. H.; Chen, Z.; Gooding, A. A.; Lin, K. J.; Kiang, C. H. Direct Observation of Multiple Pathways of Single-Stranded DNA Stretching. Phys. Rev. Lett. 2010, 105, 218104. 69. Plumridge, A.; Meisburger, S. P.; Andresen, K.; Pollack, L. The Impact of Base Stacking on the Conformations and Electrostatics of Single-Stranded DNA. Nucleic Acids Res. 2017, 45, 3932-3943. 70. Tran, T.; Childs-Disney, J. L.; Liu, B.; Guan, L.; Rzuczek, S.; Disney, M. D. Targeting the r(CGG) Repeats that Cause FXTAS with Modularly Assembled Small Molecules and Oligonucleotides. ACS Chem. Biol. 2014, 9, 904-12. 71. Disney, M. D.; Liu, B.; Yang, W. Y.; Sellier, C.; Tran, T.; Charlet-Berguerand, N.; ChildsDisney, J. L. A Small Molecule that Targets r(CGG)(exp) and Improves Defects in Fragile XAssociated Tremor Ataxia Syndrome. ACS Chem. Biol. 2012, 7, 1711-8.

TOC Graphic


26

Page 27 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60



27


Page 28 of 36

Table 1. Summary of repeat sequences with their associated diseases and their reported in vitro secondary structures. Repeat Sequence CGG CAG CTG GAA GCG CAGG ATTCT TGGAA GGGGCC

Disease Fragile X syndrome Fragile X tremor ataxia syndrome SCA 1, 2, 3, 6, 7, 12 and 17 Huntington’s disease SCA 8, Myotonic dystrophy 1 Friedreich’s ataxia Oculopharyngeal muscular dystrophy Myotonic dystrophy 2 SCA 10 SCA 31 Amyotrophic lateral sclerosis


Secondary Structures hairpin,27 quadruplex72 hairpin27 hairpin27 triplex73 hairpin74 hairpin75 DNA unwinding element76 hairpin,33, 77 antiparallel duplex78 hairpin2, quadruplex,79 R-loop2

Page 29 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Table 2. Thermodynamic parameters derived from van’t Hoff analysis (Eq. 2) for (CTG)5,5 and (CAG)5,5 hairpins in varying NaCl with T as 25 °C. ΔG° calculated from ΔH° and ΔS° by Eq. 3. Sequence NaCl (mM) ΔG° (kcal/mol) ΔH° (kcal/mol) ΔS° (cal/mol/K) (CTG)5,5

(CAG)5,5

25

-0.79 ± 0.03

-18.33 ± 0.42

-58.84 ± 1.37

100

-2.08 ± 0.24

-8.87 ± 0.80

-22.74 ± 1.59

500

-3.04 ± 0.30

-2.48 ± 0.10

1.89 ± 0.17

25

2.57 ± 0.25

-10.67 ± 0.82

-44.39 ± 2.73

100

0.88 ± 0.07

-9.42 ± 0.37

-34.55 ± 2.48

500

-0.50 ± 0.05

-6.24 ± 0.59

-19.23 ± 0.93



Page 30 of 36

Table 3. Transition-state thermodynamic parameters for (CTG)5,5 and (CAG)5,5 hairpins in the indicated NaCl conditions. Opening Transition

Closing Transition

Sequence

NaCl (mM)

ΔH ǂ (kcal/mol)

ΔS ǂ (cal/mol/K)

ΔH ǂ (kcal/mol)

ΔS ǂ (cal/mol/K)

(CTG)5,5

25

24.02 ± 0.74

51.11 ± 2.43

5.69 ± 0.37

-7.74 ± 0.60

100

20.52 ± 0.93

35.36 ± 1.02

11.65 ± 0.25

12.61 ± 0.82

500

15.87 ± 1.16

16.97 ± 0.38

13.39 ± 0.53

18.85 ± 0.73

25

13.96 ± 0.51

23.83 ± 1.69

3.29 ± 0.21

-20.57 ± 1.37

100

13.66 ± 0.75

20.66 ± 1.62

4.23 ± 0.22

-13.89 ± 1.15

500

11.74 ± 0.36

12.60 ± 1.17

5.50 ± 0.26

-6.63 ± 0.34

(CAG)5,5


Page 31 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 1. Hairpin design and results for single-molecule FRET experiments. a) Overview of hairpin states in which open conformation corresponds to the low FRET state and closed conformation corresponds to the high FRET state. b) Construct design for trinucleotide repeat-rich hairpins. Each construct contains two trinucleotide repeat domains with a specific number of CTG or CAG repeats in each domain. An A15 linker connects the trinucleotide domains. The Cy3 fluorophore was covalently attached to the 3’ end of the hairpin strand, and an internal Cy5 was positioned 7 nt from the 5’ end of the biotinylated tethering strand. Representative FRET time traces from individual molecules and cumulative smFRET histograms for symmetric CTG hairpins of increasing repeat number: c) (CTG)3,3, d) (CTG)4,4, e) (CTG)5,5, and f) (CTG)6,6 at 25 mM NaCl.



Page 32 of 36

Figure 2. Effect of stem asymmetry on the conformational behavior of (CTG)N1,N2 hairpins. smFRET histograms are shown for a) (CTG)4,4, b) (CTG)5,5, and c) (CTG)5,4 in 100 mM NaCl. Each histogram is composed from a minimum of 500 molecules. The red lines correspond to the best-fit, two-peak Gaussian distributions to the low and high FRET states. All fits had R2 values exceeding 0.96.

a)

b)


c)

Page 33 of 36

Figure 3. Effect of NaCl concentration on (CTG)5,5 hairpin behavior. a) smFRET heat map showing the normalized FRET populations as a function of increasing NaCl concentration (0, 10, 25, 50, 75, 100, 125, 250, 500, and 1000 mM). b) Measured rate constants for the hairpin opening (●) and closing (○) transitions versus total sodium concentration. c) Natural log plot of equilibrium constant (Keq) as a function of total sodium concentration. The slope of the linear fit, ΔΓ = 1.88, represents the net change in number of bound sodium ions with conformational change.

a)

c)

b)

FRET

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


0

[NaCl], M 1



Figure 4. Dependence of (CAG)5,5 hairpin behavior on NaCl concentration. a) smFRET heat map showing the normalized FRET population as a function of NaCl concentration (0, 10, 25, 50, 75, 100, 125, 250, 500, and 1000 mM). b) Measured rate constants for the opening (●) and closing (○) transitions in varying sodium concentration. c) Natural log plot of equilibrium constant (Keq) as a function of total sodium concentration. The slope of the linear fit, ΔΓ = 1.94.

a)

c)

b)

FRET

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 34 of 36

0

[NaCl], M 1


Page 35 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 5. Temperature-based smFRET measurements for (CTG)5,5 in varying NaCl. a) van’t Hoff plots showing –Rln(Keq) as a function of the inverse temperature in 25 mM (□), 100 mM (●), and 500 mM (○) NaCl. Eyring plots for the b) opening and c) closing transition rates as functions of the inverse temperature in varying NaCl with the same symbols as a). The standard-state thermodynamic parameters in Table 1 are derived from linear fits to a) using Eq. 2. The transitionstate parameters in Table 2 are derived from linear fits to b) and c) using Eqs. 4 and 5, respectively.

a)

b)


c)


Figure 6. Energy diagrams showing sequence and salt dependence for folding of trinucleotide repeat hairpins. For the (CTG)5,5 hairpin, the profiles along the reaction coordinate from the open (O) to the closed hairpin (HP) are shown for the a) free energy as well as the b) enthalpic and c) entropic components at 25 mM NaCl (red) and 500 mM NaCl (blue). For the (CAG)5,5 hairpin, the profiles along the reaction coordinate from the open (O) to the hairpin (HP) are shown for the d) free energy as well as the e) enthalpy and f) entropy components at 25 mM NaCl (red) and 500 mM NaCl (blue). All of the profiles are plotted with the HP state defined as the zero energy level.


Page 36 of 36

Sequence-Dependent Effects of Monovalent Cations on the Structural

Recommend Documents