Folding energy landscape of the thiaminepyrophosphate riboswitch aptamerPeter C. Anthonya, Christian F. Perezb, Cuauhtémoc García-Garcíac,d, and Steven M. Blockc,e,1

aBiophysics Program, bDepartment of Physics, cDepartment of Biology, dDepartment of Iberian and Latin American Cultures, and eDepartment of AppliedPhysics, Stanford University, Stanford, CA 94305

Edited by Kiyoshi Mizuuchi, National Institute of Diabetes and Digestive and Kidney Diseases, Bethesda, MD, and approved November 10, 2011 (received forreview September 13, 2011)

Riboswitches are motifs in the untranslated regions (UTRs) of RNAtranscripts that sense metabolite levels and modulate the expres-sion of the corresponding genes for metabolite import, export,synthesis, or degradation. All riboswitches contain an aptamer: anRNA structure that, upon binding ligand, folds to expose or seques-ter regulatory elements in the adjacent sequence through alterna-tive nucleotide pairing. The coupling between ligand binding andaptamer folding is central to the regulatory mechanisms of thia-mine pyrophosphate (TPP) riboswitches and has not been fullycharacterized. Here, we show that TPP aptamer folding can be de-composed into ligand-independent and -dependent steps thatcorrespond to the formation of secondary and tertiary structures,respectively. We reconstructed the full energy landscape for fold-ing of the wild-type (WT) aptamer and measured perturbations ofthis landscape arising from mutations or ligand binding. We showthat TPP binding proceeds in two steps, from aweakly to a stronglybound state. Our data imply a hierarchical folding sequence, andprovide a framework for understanding molecular mechanismthroughout the TPP riboswitch family.

optical trapping ∣ optical tweezers ∣ single molecule ∣single-molecule biophysics

Riboswitches that sense the essential coenzyme thiamine pyr-ophosphate (TPP) are found in all kingdoms of life, and reg-

ulate thiamine synthesis at the level of transcription, translation,or splicing (1–3). Members of the TPP riboswitch family sharesequence elements, architectures, and modes of ligand binding(4–9). The TPP-binding aptamer in the 3′ UTR of the thiC genefrom Arabidopsis thaliana possesses a “tuning-fork” architecture(10) comprising two sensor helix arms (P2∕3 and P4∕5) and aswitch helix (P1), all stemming from a central junction (J2∕4)(Fig. 1A). The aptamer is thought to bind its ligand as the sensorhelix arms are brought together, and bulges (J2∕3 and J4∕5) inthe arms join to form a bipartite binding pocket. Purine ribos-witches also resemble tuning forks (11, 12), but in contrast havea single binding pocket comprised largely of nucleotides from thecentral junction.

It has been proposed that riboswitches may be sorted into twofunctional types (13): Type I and Type II, of which the purine andTPP riboswitches are prototypic examples, respectively [as moreriboswitch sequences and structures have been determined, addi-tional classification schemes have been discussed (1, 14)]. Thetwo types are distinguished by binding pocket architecture andthe scale of the structural rearrangement accompanying ligandbinding, with Type I and Type II undergoing local and long-distance rearrangements, respectively. An earlier single-moleculestudy (15) of the Type I pbuE aptamer from Bacillus subtilis, whichbinds adenine, revealed that secondary and tertiary structureformation were interleaved during folding, in that a competentbinding site (constituting a tertiary element) was formed prior tothe closure of the base of the switch helix, P1 (a secondary ele-ment). Here, we have extended the single-molecule approachesused previously to investigate the folding and energetics of theTPP aptamer, which is significantly larger than the adenine

aptamer (111 nt vs. 62 nt). By measuring the energy landscapeof the TPP aptamer and its sensitivity to ligand binding andmutations, we explored the hierarchy of folding in this Type IIaptamer and obtained evidence supporting the Type I/Type IIdichotomy.

We characterized the thermodynamics and kinetics of foldingfor the TPP aptamer using a single-molecule optical-trappingassay in which a controlled force was applied via the 5′ and 3′ends of the RNA (Fig. 1B). In the absence of ligand, the aptamerunfolded through a series of transitions observed in both non-equilibrium (Fig. 1C) and equilibrium (Fig. 1D) measurements.Similar transitions were observed in RNA constructs containingonly portions of the aptamer sequence (Fig. 1A, Fig. S1, Table S1).Constant-force measurements carried out at equilibrium supplieddata for the energy, rate, and distance associated with each foldingtransition. The data from the constructs were first used to assigneach transition to a given helix, and then to reconstruct the energylandscape for secondary structure formation (Fig. 2, red). To studyaspects of tertiary structure formation, we compared aptamerfolding for wild-type and mutant sequences in the presence andabsence of ligands.

Results and DiscussionWe observed and measured individual folding transitions withinsingle WTaptamer molecules by holding each molecule at equi-librium, and gradually lowering the applied force while recordingthe molecular end-to-end extension (Fig. 1D). Our measure-ments allowed us to identify each transition and calculate itsrespective contribution to the folding energy landscape (Fig. 2).In the absence of ligand, the energetic stabilities of the helicesdecreased systematically, going from the distal ends of the sensorhelix arms toward the 5′ and 3′ termini of the RNA. Previousmodels have suggested that aptamer folding, in concert with TPPbinding, proceeds with a similar “outside-in” directionality: TPPbinding juxtaposes the sensor helices, promoting formation ofthe central junction and thereby stabilizing the switch helix(10, 16). Our data demonstrate that TPP binding is not requiredto fold P1 (the least stable helix), but is not inconsistent withmechanisms where this switch helix, in the context of a completeriboswitch, remains unfolded in the absence of TPP due to alter-native pairing with the expression platform (17–19).

The significant energy differences between most folding tran-sitions (Fig. 2) may serve to prevent misfolding of the aptamer,which we never observed, and the relatively high stabilities ofthe P3 and P5 helices likely ensure correct nucleation of theTPP binding sites at J2∕3 and J4∕5. In particular, P3 is the firsthelix of the aptamer to be transcribed, and the ∼32 nt at its distal

Author contributions: P.C.A., C.F.P., C.G.-G., and S.M.B. designed research; P.C.A., C.F.P., andC.G.-G. performed research; P.C.A. analyzed data; and P.C.A. and S.M.B. wrote the paper.

The authors declare no conflict of interest.

This article is a PNAS Direct Submission.1To whom correspondence should be addressed. E-mail: sblock@stanford.edu.

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1115045109/-/DCSupplemental.

www.pnas.org/cgi/doi/10.1073/pnas.1115045109 PNAS ∣ January 31, 2012 ∣ vol. 109 ∣ no. 5 ∣ 1485–1489

BIOPH

YSICSAND

COMPU

TATIONALBIOLO

GY

Dow

nloa

ded

by g

uest

on

July

20,

202

1

end, which are not required for TPP binding or L5-P3 docking(10, 20), participated in the highest-force folding transition inthe aptamer (P3high). The P3high transition occurred separatelyfrom that involving the ∼12 nt at the base of the helix (P3low)(Table S1). The size of the P3 element varies greatly amongspecies (6, 9), and its distal portion may have evolved as a “ther-modynamic anchor” for the aptamer, ensuring correct folding

by suppressing competing RNA structures incompatible with theriboswitch mechanism.

We determined the free energy change for complete unfoldingof the bare aptamer (61� 3 kcal∕mol) by measuring distributionsof the work done to unfold and refold the aptamer under none-quilibrium conditions (Fig. 1C, inset). This value closely matchedthe sum of energies to unfold the individual helices (66� 3 kcal∕mol), suggesting that no strong tertiary interactions are presentin the absence of TPP. Both these experimental values are some-what greater than that estimated by the mfold program(∼53 kcal∕mol) (21). The difference is likely attributable toMg2þ, which is not accounted for by mfold, but was present at4 mM here. Previous structural studies using small-angle X-rayscattering (SAXS) have found that under similar conditions,the sensor helix arms are splayed apart (20), but the aptamer be-comes partially compacted as magnesium levels are increased(20, 22). In light of our findings, this compactness may be dueto divalent cations stabilizing these helices without participatingin energetically significant contacts between them.

Whether the TPP aptamer is “preorganized,” that is, whetherit contains secondary or tertiary structure to facilitate ligand bind-ing, has garnered controversy. Upon solving the crystal structureof the TPP-bound aptamer from A. thaliana, Thore, et al. conjec-tured that J2∕4 and P1 do not form until TPP binds and bridgesthe two sensor helices (10). Lang, et al. presented kinetic datain support of this conjecture for the shorter Escherichia coli thiMaptamer, but could not rule out the possibility that the P1 helixwas paired prior to TPP binding (16). In a calorimetric study,Kulshina, et al. observed diminishing enthalpic gains and entropicpenalties for TPP binding to the same aptamer as the magnesiumconcentration was raised, and therefore suggested that magne-sium induces preorganization (5). However, the authors did notdemonstrate that this preorganization included any tertiary struc-ture. Most recently, Steen, et al. chemically modified the thiMRNA in the absence of ligand and observed a pattern of pro-tections indicative of fully formed secondary structure, but no

P1P2

2 kb DNARNA

laser beam

dig/anti-dig

~35 bp RNA/DNA hybrids

6 nt spacers

laser beam

1 kb DNA

biotin/avidin

B

200

150

100

50

0

Cou

nts

6004002000Work (pN·nm)

Extension:increasingdecreasing

40

30

20

10

0

For

ce (

pN)

11001000900800Extension (nm)

10 n

m

2 s

}}}

P5

P3high

P4

F

UC D

A

P3

J4/5

P4

P5

J2/4P1

P2/3

P3P3top

P4/5

1 kb dsDNA handle

2 kb dsDNA handle

A. thaliana thiC TPP riboswitch aptamer

3'

5'

Fig. 1. Single-molecule observations of TPP riboswitch aptamer folding.(A) Sequence and secondary structure of the A. thaliana thiC TPP-bindingaptamer, indicating mutations (red) and construct sequences measured inisolation (boxed). Short spacer sequences (lower case letters, gray) wereused to connect all constructs to dsDNA handles. Nucleotides involved inthe P3low and P3high transitions are indicated (dashed line). (B) Experimen-tal geometry of the “dumbbell” optical-trapping assay with componentslabeled (not to scale). (C) Representative nonequilibrium FECs for the WTaptamer in the absence of ligand, obtained by increasing (red) or decreasing(black) the distance between traps (100 cycles). WLC curve-fits (blue) to thedata below or above all transitions are shown. Inset, histograms of the un-folding and refolding work (5 molecules, 1,291 cycles). (D) Left, equilibriumtraces of extension acquired at constant force showing individual foldingtransitions; the corresponding transitions in the FECs are indicated (blackarrows). Right, histograms of extension (red) and Gaussian fits (black). Thefully folded (F) and unfolded (U) states, as well as the pairs of peaks corre-sponding to specific transitions (P1, P2, P3high, P4, P5) are labeled.

Fre

e en

ergy

Extension

WT

A90G

A105G

F P1

P3low

P3high(U)

P2

P4P5

5 nm

10 k

cal/m

ol

WTA90G,

2 mM TPPWT,

2 mM TWT,

2 mM TMPWT,

2 mM TPP

F

P1

Fig. 2. Reconstructed energy landscapes for thiC aptamers. The experimen-tally determined free energies for folded, unfolded, and intermediate statesare plotted vs. extension, along with the transition energy barriers separat-ing these for the WT, A90G, and A105G aptamers (mean� s:e:; see SIMethods). States are labeled in the order that unfolding transitions occurredas the extension was raised (e.g., P1 is unfolded in the state marked “P1”);smooth curves (red, blue, green) connect the points. The landscapes for theWT (red filled circles) and mutant aptamers (blue and green filled circles)were indistinguishable except where indicated. Landscapes are shown tiltedby the work done by 8 pN applied force. Inset, view of the F-to-P1 transitionillustrating the stabilization of the F state by ligand binding under the indi-cated conditions (legend).

1486 ∣ www.pnas.org/cgi/doi/10.1073/pnas.1115045109 Anthony et al.

Dow

nloa

ded

by g

uest

on

July

20,

202

1

tertiary structure (23). Our data are consistent with all the data(but not necessarily all the interpretations) in these prior studies,as well as with the SAXS studies mentioned above, and argue thatany preorganization of the A. thaliana thiC aptamer involves theformation of helices P1-P5 and the juxtaposition of those helicesdue to stacking or primary connectivity.

In the presence of saturating TPP, larger, abrupt transitions(“rips”) were observed in force-extension curves (FECs) as theaptamer was unfolded (Fig. 3A). Worm-like chain (WLC) func-tions fitted to the appropriate portions of each curve revealedthat 109� 6 nt are released in these rips, which equals the num-ber measured upon unfolding all helices in the absence of TPP(Fig. 1C, Table S2) and the total length of the aptamer. Thisequivalence suggests that the aptamer unfolds monolithicallyupon TPP dissociation when force is increased slowly. As theaptamer was gradually refolded near equilibrium in the presenceof TPP, the folding behaviors of the P3, P4, and P5 helices wereunperturbed (Table S1), but the free, reversible folding of the P2and P1 helices was interrupted by an event, presumed to be TPPbinding, that locked the aptamer into its fully folded state(Fig. 3A). Refolding FECs appeared identical regardless ofwhether TPP was present (Fig. 1C, Fig. 3A). All together, theseobservations indicate that TPP binds concomitantly with P1 fold-ing. Analyses of the rip-force distributions were carried out usingthe methods developed by Dudko (24) and Maitra (25), whichsupplied the heights and locations of the transition energy bar-riers associated with the rips, and placed the barrier for TPP dis-sociation at a distance of 12� 1 nt from the fully folded state ofthe aptamer (Table S2). This distance corresponds closely to thelength of the P1 helix.

We determined the additional energetic stability conferred byTPP binding, including any tertiary contacts formed, by compar-ing the free energies for aptamer unfolding in the presence andabsence of TPP, computed from the work distributions (Fig. 1C,Table S2, Fig. S2). The difference, 17� 5 kcal∕mol, is compar-able to the value of 11� 1 kcal∕mol measured calorimetricallyfor the thiM aptamer from E. coli (5). Because ligand bindingis closely correlated with the folding of the P1 helix in our experi-ment, we interpreted this difference to correspond to the addi-tional stabilization imparted to the fully folded state of theaptamer (Fig. 2, inset).

This additional stabilization could be abolished by mutatingthe conserved adenine residue A105 in the central junction to

guanine, corroborating findings from previous studies (4, 18).When present at saturating concentrations, TPP bound theA105G aptamer in just 2% of repeated nonequilibrium foldingcycles (Fig. 3B), compared with 98% for the WT. Moreover, thefolding behavior of the A105G-P2 and P1 helices at equilibriumwas unaffected by TPP levels (Table S1). However, P1 folded fas-ter in the A105G aptamer than WT in the absence of TPP, andinvolved 6� 2 fewer nucleotides, consistent with the P1 helixfraying, due to a shift in the registration of the 5′ and 3′ strands(Fig. S3). Such a misregistration is, in fact, predicted by mfold,and serves to explain the lower affinity of the mutant for TPP.In the WTaptamer, the A105 residue is unpaired and participatesin several tertiary contacts across J2∕4 (10). These contacts wouldbe broken were this residue to pair in a misregistered P1 helix,disfavoring the orientation of the sensor helices required to bindTPP.

In addition to tertiary contacts formed in J2∕4, tertiary con-tacts form between the L5 loop and P3 helix when TPP is bound(10). Mutation of certain L5 residues prevents TPP-inducedcompaction in the E. coli aptamer (5). Here, we observed thatmutating A90 to G reduced the TPP-induced stabilization ofthe fully folded state by ∼30%, to 12� 5 kcal∕mol (Fig. 2, inset;Fig. 3C). The reconstructed energy landscapes for the WT andA90G aptamers did not differ significantly in the absence ofTPP (the sum of A90G helix energies is 66� 3 kcal∕mol;Fig. S4), but in the presence of TPP and near the equilibriumfolding forces for the P2 and P1 helices, ligand bound to theA90G aptamer only transiently (Fig. 3C), dissociating on a timescale of seconds, rather than hours.

We also acquired FECs (Fig. S2) and constant-force data(Fig. S5) for the WTaptamer binding to the substrates thiamine(T) and thiamine monophosphate (TMP). These alternativeligands carry fewer phosphates than TPP and are unable to bindthe P4∕5 arm via native, magnesium-mediated contacts (19, 26–28). Nevertheless, the WT aptamer displayed similar energeticstabilization by either Tor TMP, when present at saturating con-centrations, as did the A90G aptamer bound to TPP (Fig. 2, inset;Table S2). In addition, the WT aptamer also bound T or TMPtransiently as P1 folded at equilibrium (Fig. S5). These character-istics support the existence of a common, weak-binding statefor the aptamer that either lacks full coordination of TPP at itsbinding sites, or lacks productive L5-P3 docking around TPP.

40

30

20

10

0

For

ce (

pN)A

B

C

WT2 mM TPP

A105G2 mM TPP

9.0 pN

8.2 pN

5 s

10 nm

Time

Ext

.

A90G2 mM TPP

9.2 pN

40

30

20

10

0

For

ce (

pN)

40

30

20

10

0

For

ce (

pN)

11001000900800Extension (nm)

Fig. 3. Effects of mutations on TPP binding. Left, records of extension vs. time for P1 and P2 folding under constant forces (as indicated) for aptamers in thepresence of TPP, showing presumed TPP binding events (black arrows): (A) WT plus 2 μM TPP, (B) A105G plus 2 mM TPP, and (C) A90G plus 2 mM TPP. Right, FECsof WT, A105G, and A90G aptamers (150 cycles each) in the presence of 2 mM TPP, showing transitions (red, unfolding; black, refolding).

Anthony et al. PNAS ∣ January 31, 2012 ∣ vol. 109 ∣ no. 5 ∣ 1487

BIOPH

YSICSAND

COMPU

TATIONALBIOLO

GY

Dow

nloa

ded

by g

uest

on

July

20,

202

1

We obtained corroborating evidence for such a state by prob-ing the kinetics of TPP binding to the WTaptamer at subsaturat-ing ligand concentrations. By unfolding each molecule under

nonequilibrium conditions and then relaxing the force rapidly,we measured the probability that refolding and subsequentTPP binding would occur after a given interval between unfoldingcycles, signaled by a TPP-dependent rip. We observed two popu-lations of rip forces (Fig. 4 A and B) with high and low meanvalues, comparable with those measured for dissociating TPPfrom the WT and A90G aptamers, respectively, at saturatingligand concentrations (Fig. 3 A and C). The probabilities forobserving a TPP-dependent rip at any force (Fig. 4C) and forobserving the rip at high force (Fig. 4D) both increased with therefolding time and TPP concentration. These probabilities wereglobally fit by a minimal, four-state kinetic model (Fig. 4D) thatcollapses all helix-folding transitions into a single irreversible, faststep and incorporates two TPP-bound states, representing weak,initial binding followed by strong, final binding. A ligand disso-ciation constant of 7� 1 × 10−8 M was calculated from the mod-el, which is consistent with previous estimates (5, 18, 19). Wepropose that progression from weak to strong binding, which gen-erally occurred when the WT aptamer was paired with TPP, butnot when a mutant RNA sequence or alternative ligand was sub-stituted, may serve as a final step in the folding process to ensureselectivity and the fidelity of riboswitch activation.

Taken all together, our single-molecule data suggest that sec-ondary structure forms quickly in the TPP riboswitch aptamer,and that it does so independent of ligand binding, whereas ter-tiary structure forms comparatively slowly, and does so in concertwith ligand binding. We note that we could only indirectly observeligand-dependent juxtaposition of the P2∕3 and P4∕5 helix arms,and L5-P3 docking, because these tertiary folding events lacka projection along our reaction coordinate, the molecular end-to-end extension (Figs. 1, 2). Therefore, we cannot strictly ruleout the possibility that some minor amount of tertiary structuremay form in the absence of ligand. Nevertheless, the hierarchicalsequence of folding events that we observed stands in stark con-trast to that of the adenine riboswitch aptamer (15), and it high-lights key differences between folding in Type I (e.g., adenine)and Type II (e.g., TPP) riboswitches (13). In our experiments,productive TPP binding generally required P1 helix folding, butunder some conditions TPP was observed briefly to remainstrongly bound if P1 (and in rare cases, P2) became unpaired(Fig. S6) during unfolding. Such unpairing will weaken or disruptJ2∕4, and therefore indicates that most of the energy for bindingligand arises from the interactions between the sensor helix arms,which TPP brings closer together.

In A. thaliana, TPP binding to the riboswitch aptamer pro-motes the formation of a longer, less stable thiC RNA, by causinga transcript processing site to be spliced out of the 3′ UTR (4, 7)(Fig. S7A). Ligand binding is thought to occur cotranscriptionally,and acts to prevent base-pairing between the 5′ splice site(located ∼150 nt upstream of the aptamer) and the P4∕5 helixarm of the aptamer (7). Our data suggest that ligand binding can-not occur until the entire aptamer has been transcribed and thefinal elements of secondary structure (J2∕4, P1) have formed.This arrangement would allow the 5′ strand of P4∕5 to pair with,and sequester, the 5′ splice site immediately after transcription,thereby preventing splicing even at high TPP concentrations(Fig. S7B). Because splicing-incompetent configurations of thenascent 3′ UTR likely do not contain functional aptamers, theymay represent kinetic traps, escape from which requires reconfi-guration of base-pairing in the 3′ UTR, coupled with TPP bind-ing. The 3′ splice site is situated within the aptamer, and itsaccessibility may change when TPP binds, as has been suggested(4). Based on our findings, this change is likely to occur as TPPbinding progresses from weak to strong.

We anticipate that our study will serve as a starting point fora more complete biophysical picture of aptamer folding coupledto TPP binding and mRNA splicing in A. thaliana, and of TPPriboswitch function in other organisms. By extending the single-

A

C

D

B

Refolding time (s)

Fig. 4. Kinetics of TPP binding. (A) Representative FECs (N ¼ 38) showingTPP-dependent rips at high force (red) or low force (green), or lacking eitherkind of rip (blue). (B) Histograms (mean� s:e:) of TPP-dependent rip forcesdetermined after 1- or 10-s refolding times (10 μM TPP; N ¼ 100 rips per re-folding time). (C) Probability (solid circles, mean� binomial error) that TPPis bound to the WTaptamer (indicated by a TPP-dependent rip) as a functionof the refolding time for the TPP concentration indicated (color legend).(D) Probability (solid circles, mean� binomial error) that TPP binding isstrong (indicated by a rip at high force) as a function of the refolding time,and a minimal (four-state) kinetic model for binding, with both weak(F′) and strong (F′′) binding states (bottom). The legend (inset) shows thebest-fit rate constants; datasets in (C) and (D) were globally fitted (solid lines;colored by TPP concentration as in (C)).

1488 ∣ www.pnas.org/cgi/doi/10.1073/pnas.1115045109 Anthony et al.

Dow

nloa

ded

by g

uest

on

July

20,

202

1

molecule methods applied previously to the folding of theadenine-binding pbuE aptamer (15), we directly measured theenergetics, kinetics, and hierarchy of folding for an aptamer thatis nearly twice as large. Several new approaches were employed tomake these measurements, including: a “divide-and-conquer”strategy to assign individual folding events to the overall aptamerstructure, the use of analytical methods from nonequilibriumthermodynamics to quantify the impact of ligand binding on thefolding energy landscape, and use of point mutations and alter-native ligands to perturb and elucidate folding. We expect single-molecule methods to be broadly applicable in future studies ofadditional riboswitches and structured RNAs.

Materials and MethodsSample Preparation. RNA constructs were transcribed in vitro and annealed to∼kb-long dsDNA handles, each having a single-stranded overhang at one endand a biotin or digoxigenin modification at the other (see SI Methods).Dumbbells (29) were assembled by incubating the resulting construct withavidin- and anti-digoxigenin-coated polystyrene beads, and measured in50 mM Hepes buffer containing (unless otherwise noted) 130 mM KCl and4 mM Mg2þ, plus trace components and ligand.

Instrument. All data were acquired on a dual-beam optical-trapping micro-scope (30, 31), in which the trapping beams were steered using acousto-optical deflectors. Bead positions were detected using duolateral position-sensitive detectors.

Measurement of Nonequilibrium FECs with Hysteresis. Global aptamer unfold-ing and refolding events were induced by repeatedly moving the traps apartand together, respectively, at rates of ∼60–200 nm∕s. Rips were identified aslocal maxima in the unfolding FECs. Parameters describing the free energy

barriers for folding transitions or ligand binding (Δx‡, ΔG‡, koff) were ex-tracted from nonequilibrium distributions of rip forces (24, 25). The workcorresponding to unfolding or refolding was calculated by integrating eachFEC and subtracting the work of stretching the handles and fully unfoldedRNA. The total free energy of unfolding (ΔG) was calculated from an analysisof work histograms (32).

Folding Energy Landscapes. Equilibrium measurements conducted at constantforce were carried out for the individual folding transitions in the full-lengthand truncated aptamer constructs. Each folding transition was analyzed as atwo-state system, and the associated parameters describing the energeticsand kinetics of folding (Δx, Δx‡, F1∕2, and k1∕2) were extracted (31). Theset of folding transitions present in each construct, and values of the para-meters describing each transition in this set, were compared among the con-structs and correlated with the underlying sequence in order to assigntransitions to specific helices. These parameters were then used as described(15) to reconstruct landscapes representing aptamer folding in the absenceof ligand. Perturbations of the energy landscapes by ligand binding near thefully folded state were characterized using the parameters obtained from theanalysis of nonequilibrium FECs.

TPP Binding Kinetics. The frequency and strength of TPP binding to the WTaptamer were determined from FECs measured while moving the traps apartat ∼400 nm∕s to induce unfolding. The TPP concentration and time betweencycles were varied. TPP-dependent rips were identified by eye and classifiedas either low-force or high-force using the P3high transition as a threshold.

ACKNOWLEDGMENTS.We thank M. Ali, A. Sim, V. Chu, V. Pande, D. Herschlag,S. Doniach, and members of the Block lab for helpful comments. This workwas supported by a National Science Foundation Graduate Research Fellow-ship (to P.C.A.) and a grant from the National Institute of GeneralMedical Sciences (to S.M.B.).

1. Roth A, Breaker RR (2009) The structural and functional diversity of metabolite-binding riboswitches. Annu Rev Biochem 78:305–334.

2. Miranda-Rios J (2007) The THI-box riboswitch, or how RNA binds thiamin pyropho-sphate. Structure 15:259–265.

3. Bocobza SE, Aharoni A (2008) Switching the light on plant riboswitches. Trends PlantSci 13:526–533.

4. Bocobza S, et al. (2007) Riboswitch-dependent gene regulation and its evolution in theplant kingdom. Genes Dev 21:2874–2879.

5. Kulshina N, Edwards TE, Ferre-D’Amare AR (2010) Thermodynamic analysis of ligandbinding and ligand binding-induced tertiary structure formation by the thiaminepyrophosphate riboswitch. RNA 16:186–196.

6. Sudarsan N, Barrick JE, Breaker RR (2003) Metabolite-binding RNA domains arepresent in the genes of eukaryotes. RNA 9:644–647.

7. Wachter A, et al. (2007) Riboswitch control of gene expression in plants by splicing andalternative 3′ end processing of mRNAs. Plant Cell 19:3437–3450.

8. Barrick JE, Breaker RR (2007) The distributions, mechanisms, and structures of meta-bolite-binding riboswitches. Genome Biol 8:R239.

9. Kubodera T, et al. (2003) Thiamine-regulated gene expression of Aspergillus oryzaethiA requires splicing of the intron containing a riboswitch-like domain in the5′-UTR. FEBS Lett 555:516–520.

10. Thore S, Leibundgut M, Ban N (2006) Structure of the eukaryotic thiamine pyropho-sphate riboswitch with its regulatory ligand. Science 312:1208–1211.

11. Batey RT, Gilbert SD, Montange RK (2004) Structure of a natural guanine-responsiveriboswitch complexed with the metabolite hypoxanthine. Nature 432:411–415.

12. Serganov A, et al. (2004) Structural basis for discriminative regulation of gene expres-sion by adenine- and guanine-sensing mRNAs. Chem Biol 11:1729–1741.

13. Montange RK, Batey RT (2008) Riboswitches: emerging themes in RNA structure andfunction. Annu Rev Biophys Bio 37:117–133.

14. Garst AD, Batey RT, Edwards AL (2011) Riboswitches: structures and mechanisms. ColdSpring Harb Perspect Biol 3:a003533.

15. Greenleaf WJ, Frieda KL, Foster DA, Woodside MT, Block SM (2008) Direct observationof hierarchical folding in single riboswitch aptamers. Science 319:630–633.

16. Lang K, Rieder R, Micura R (2007) Ligand-induced folding of the thiM TPP riboswitchinvestigated by a structure-based fluorescence spectroscopic approach. Nucleic AcidsRes 35:5370–5378.

17. CroftMT,MoulinM,WebbME, Smith AG (2007) Thiamine biosynthesis in algae is regu-lated by riboswitches. Proc Natl Acad Sci USA 104:20770–20775.

18. Sudarsan N, Cohen-Chalamish S, Nakamura S, Emilsson GM, Breaker RR (2005) Thia-mine pyrophosphate riboswitches are targets for the antimicrobial compoundpyrithiamine. Chem Biol 12:1325–1335.

19. Winkler W, Nahvi A, Breaker RR (2002) Thiamine derivatives bind messenger RNAsdirectly to regulate bacterial gene expression. Nature 419:952–956.

20. Ali M, Lipfert J, Seifert S, Herschlag D, Doniach S (2010) The ligand-free state of the TPPriboswitch: a partially folded RNA structure. J Mol Biol 396:153–165.

21. ZukerM (2003) Mfold web server for nucleic acid folding and hybridization prediction.Nucleic Acids Res 31:3406–3415.

22. Baird NJ, Ferre-D’Amare AR (2010) Idiosyncratically tuned switching behavior ofriboswitch aptamer domains revealed by comparative small-angle X-ray scatteringanalysis. RNA 16:598–609.

23. Steen KA, Malhotra A, Weeks KM (2010) Selective 2′-hydroxyl acylation analyzed byprotection from exoribonuclease. J Am Chem Soc 132:9940–9943.

24. Dudko OK, Hummer G, Szabo A (2006) Intrinsic rates and activation free energies fromsingle-molecule pulling experiments. Phys Rev Lett 96:108101.

25. Maitra A, Arya G (2011) Influence of pulling handles and device stiffness in single-molecule force spectroscopy. Phys Chem Chem Phys 13:1836–1842.

26. Edwards TE, Ferre-D’Amare AR (2006) Crystal structures of the thi-box riboswitchbound to thiamine pyrophosphate analogs reveal adaptive RNA-small moleculerecognition. Structure 14:1459–1468.

27. Noeske J, Richter C, Stirnal E, Schwalbe H, Wohnert J (2006) Phosphate-group recogni-tion by the aptamer domain of the thiamine pyrophosphate sensing riboswitch.Chembiochem 7:1451–1456.

28. Thore S, Frick C, Ban N (2008) Structural basis of thiamine pyrophosphate analoguesbinding to the eukaryotic riboswitch. J Am Chem Soc 130:8116–8117.

29. Woodside MT, et al. (2006) Direct measurement of the full, sequence-dependent fold-ing landscape of a nucleic acid. Science 314:1001–1004.

30. Neuman KC, Block SM (2004) Optical trapping. Rev Sci Instrum 75:2787–2809.31. Woodside MT, et al. (2006) Nanomechanical measurements of the sequence-

dependent folding landscapes of single nucleic acid hairpins. Proc Natl Acad SciUSA 103:6190–6195.

32. Shirts MR, Bair E, Hooker G, Pande VS (2003) Equilibrium free energies from nonequi-librium measurements using maximum-likelihood methods. Phys Rev Lett 91:140601.

Anthony et al. PNAS ∣ January 31, 2012 ∣ vol. 109 ∣ no. 5 ∣ 1489

BIOPH

YSICSAND

COMPU

TATIONALBIOLO

GY

Dow

nloa

ded

by g

uest

on

July

20,

202

1