ANRV345-BI77-10 ARI 28 April 2008 11:59

Recent Advancesin Optical TweezersJeffrey R. Moffitt,1 Yann R. Chemla,3

Steven B. Smith,1 and Carlos Bustamante1,2

1Department of Physics, 2Departments of Chemistry, and Molecular and Cell Biologyand Howard Hughes Medical Institute, University of California, Berkeley,California 94720; email: carlos@alice.berkeley.edu3Department of Physics, University of Illinois, Urbana-Champaign, Urbana,Illinois 61801

Annu. Rev. Biochem. 2008. 77:205–28

First published online as a Review in Advance onFebruary 28, 2008

The Annual Review of Biochemistry is online atbiochem.annualreviews.org

This article’s doi:10.1146/annurev.biochem.77.043007.090225

Copyright c© 2008 by Annual Reviews.All rights reserved

0066-4154/08/0707-0205$20.00

Key Words

force, hybrid-instruments, molecular motors, single molecule,torque

AbstractIt has been over 20 years since the pioneering work of Arthur Ashkin,and in the intervening years, the field of optical tweezers has growntremendously. Optical tweezers are now being used in the inves-tigation of an increasing number of biochemical and biophysicalprocesses, from the basic mechanical properties of biological poly-mers to the multitude of molecular machines that drive the internaldynamics of the cell. Innovation, however, continues in all areas ofinstrumentation and technique, with much of this work focusing onthe refinement of established methods and on the integration of thistool with other forms of single-molecule manipulation or detection.Although technical in nature, these developments have importantimplications for the expanded use of optical tweezers in biochemi-cal research and thus should be of general interest. In this review,we address these recent advances and speculate on possible futuredevelopments.

205

AR FurtherClick here for quick links to Annual Reviews content online, including:

• Other articles in this volume• Top cited articles• Top downloaded articles• AR’s comprehensive search

Ann

u. R

ev. B

ioch

em. 2

008.

77:2

05-2

28. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Bol

ogna

on

06/2

7/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-10 ARI 28 April 2008 11:59

Contents

PROLOGUE . . . . . . . . . . . . . . . . . . . . . . . 206INTRODUCTION TO

OPTICAL TWEEZERS . . . . . . . . . 207How Do You Trap? . . . . . . . . . . . . . . . 207How Do You Manipulate? . . . . . . . . 208How Do You Detect? . . . . . . . . . . . . . 208How Do You Measure? . . . . . . . . . . . 209

IMPROVEMENTS INRESOLUTION . . . . . . . . . . . . . . . . . 209Why Do You Want Better

Resolution? . . . . . . . . . . . . . . . . . . . 209What Limits Resolution? . . . . . . . . . 210How Do You Remove

Experimental Noise? . . . . . . . . . . 211How Do You Address

Brownian Noise? . . . . . . . . . . . . . . 212Are Two Traps Better than One? . . 213What Resolution Has Been

Achieved? . . . . . . . . . . . . . . . . . . . . . 214What Is Needed

to Observe Steps? . . . . . . . . . . . . . 214What about Accuracy? . . . . . . . . . . . . 215

HYBRID AND NOVELINSTRUMENTS . . . . . . . . . . . . . . . 215

Why Create Hybridand Novel Instruments? . . . . . . . 216

Why Manipulate Morethan One Molecule? . . . . . . . . . . . 216

How Do You Create NovelOptical Potentials? . . . . . . . . . . . . 216

Why Add Torque to an OpticalTweezer? . . . . . . . . . . . . . . . . . . . . . . 218

How Do You Add RotationalControl to OpticalTweezers? . . . . . . . . . . . . . . . . . . . . . 218

What Other forms of ManipulationCan Be Integrated with OpticalTweezers? . . . . . . . . . . . . . . . . . . . . . 220

How is Single-MoleculeFluorescence Used? . . . . . . . . . . . 221

How Do You Use Fluorescenceto Visualize Moleculesin Optical Tweezers?. . . . . . . . . . . 221

Can You Use Single Fluorophoreswith Optical Tweezers? . . . . . . . . 222

CONCLUSIONS ANDPERSPECTIVES . . . . . . . . . . . . . . . . 223

PROLOGUE

Light carries both linear and angular momen-tum and can thus exert forces and torques onmatter. Optical tweezers exploit this funda-mental property to trap objects in a poten-tial well formed by light (1, 2). This tech-nique allows the manipulation of microscopicobjects in applications ranging from particlesorting (3) to microfabrication (4, 5). More-over, optical tweezers can be used as quan-titative tools to exert calibrated forces onsystems of interest as well as accurately andsensitively measure the forces and displace-ments generated by these systems. Such an-alytical optical tweezers have been used in avariety of applications ranging from mi-crorheology (6) to the study of colloidalhydrodynamics (7–10) and nonequilibriumthermodynamics (11–16). More importantly,

analytical optical tweezers have developedinto a powerful tool in molecular biology,biochemistry, and biophysics, where they areused to manipulate and interrogate individ-ual molecules. From these studies, scientistsare gaining essential new insights into themechanical properties of biological macro-molecules and the dynamics and mechanismsof molecular motors—the proteins and com-plexes of proteins that use chemical energy toperform mechanical tasks in the cell.

In the past decade and a half, it has becomeincreasingly evident that force is involved inmany facets of cellular life, ranging from theobvious—the transport of cellular cargo bymotors, such as myosin, kinesin, and dynein—to the more subtle and speculative, such asthe strain induced on an enzyme and its sub-strate during catalysis (17). With the ability to

206 Moffitt et al.

Ann

u. R

ev. B

ioch

em. 2

008.

77:2

05-2

28. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Bol

ogna

on

06/2

7/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-10 ARI 28 April 2008 11:59

apply forces on individual molecules and mea-sure the forces generated in their chemicalreactions, analytical optical tweezers are ide-ally suited to investigate such mechanochem-ical transformations. However, despite theirunique contributions to the study of single-molecule processes, these techniques have hadlimitations. First, they have lacked the sensi-tivity to study many of these molecular pro-cesses at their fundamental spatiotemporalscales—the angstrom to nanometer distancescales and millisecond to second timescales.Second, studies using these methods havelargely been limited to simple biological sys-tems, consisting only of a few componentsobserved in isolation, far from their nativecellular context. It has become clear that tocapture the full complexity of the cell andits components, a new experimental toolsetwill be needed, capable of manipulating morecomplicated biological systems and measur-ing their conformation changes with greatersensitivity.

In recent years, parallel advances in op-tical trapping methodologies have aimed ataddressing these limitations, on one handimproving the resolution and accuracy of cur-rent instrumentation, and on the other devel-oping new hybrid instruments, which com-bine various single-molecule techniques forthe manipulation of complex systems and thesimultaneous measurement of multiple ob-servables. In this review, we discuss this nextgeneration of instrumentation and its poten-tial for deepening our understanding of fun-damental biological processes. Because thesedevelopments are inherently technical, thisreview requires some knowledge of opticaltweezers: the principles by which they func-tion and the basic instrumentation requiredto build them. To assist readers new to thistopic, we start with a brief introduction to op-tical trapping, and throughout we refer theinterested reader to the excellent literatureon this subject. Following this introduction,we discuss the recent technical and theoreti-cal advances that have refined the resolution ofanalytical optical tweezers, culminating in the

ability to detect, for the first time, motions ofbiological systems as small as a few angstromson the one-second timescale, with accuracyapproaching a few percent. In parallel to thisdevelopment, there has been significant effortto generalize the traditional optical tweezersby incorporating them with other forms ofsingle-molecule manipulation and detection.Thus, in the third section of this review, wediscuss this current work and the biologicalproblems that have motivated these develop-ments. Finally, we conclude with some specu-lation on what the future holds for the devel-opment and application of optical tweezers inthe study of complex biological systems.

INTRODUCTION TOOPTICAL TWEEZERS

The physical principles behind optical trap-ping are relatively simple and straightforwardas are the instrumentation techniques neces-sary to build a standard optical tweezers. Inthis introductory section, we briefly reviewthe basic principles of optical trapping to aidthose new to the field and throughout refer toadditional reading in the literature. Readersfamiliar with this topic may want to skip thissection.

How Do You Trap?

Optical traps involve the balance of two typesof optical forces: scattering forces, which pushobjects along the direction of propagation ofthe light, and gradient forces, which pull ob-jects along the spatial gradient of light inten-sity (2). When gradient optical forces exceedthose from scattering, an object is attractedto the point of highest intensity formed byfocused light and can be stably trapped atthis position in all three dimensions (2). Inpractical implementations of an optical trap,a near-infrared laser beam is tightly focused bya high-numerical aperture microscope objec-tive to create the large spatial gradient in lightintensity necessary to form a stable trap (18).In the vicinity of this focus, the optical trap

www.annualreviews.org • Recent Advances in Optical Tweezers 207

Ann

u. R

ev. B

ioch

em. 2

008.

77:2

05-2

28. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Bol

ogna

on

06/2

7/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-10 ARI 28 April 2008 11:59

ΔxΔx Δx Δx

a b c

Figure 1Different experimental geometries for optical tweezers. (a) For processive cytoskeletal motors, such askinesin, the motor is typically attached directly to a polystyrene bead held in an optical trap, and thefilament is attached to the surface of a sample chamber. Motions of the motor are revealed by motions ofthe trapped bead. (b) It is also possible to attach one end of the biological system to a second polystyrenebead suctioned onto the end of a micropipette. The motion of the biological system, such as theunfolding of a RNA hairpin, is again revealed in the motion of the trapped bead. (c) The second bead canalso be held in a second optical trap. In this case, changes in the length of the tethered DNA by theaction of a bacteriophage portal motor are revealed in the motions of both beads. The relative motion ofeach bead depends on the relative stiffness of the two optical traps.

Back-focal-planeinterferometry:detection methodthat uses theinterference patternformed between thetrapping laser andthe scattered light tomeasure beadmovements

behaves as a linear “Hookean” spring, gener-ating forces on an object proportional to itsdisplacement from the center of the trap.

How Do You Manipulate?

The magnitude of these optical forces isgenerally insufficient to stably trap biologicalmacromolecules themselves, but more thanadequate for manipulating microscopicdielectric objects,1 such as micron-sizedpolystyrene beads, which can be biochem-ically linked to these molecules of interest.In the majority of single-molecule opticaltweezers experiments, these beads serve ashandles to a biological system, allowing itsmanipulation inside a sample chamber. Exert-ing calibrated forces on this molecule usuallyrequires that the biological system is also at-tached at the other end. Typically this secondattachment point is the surface of a samplechamber (Figure 1a), a second bead held atopa micropipette by suction (Figure 1b), or abead held in a second optical trap (Figure 1c)(19–21). In this fashion, the system can bestretched by moving the optical trap relative

1Typically, optical traps can generate forces of several pi-coNewtons (pN), or 1 × 10−12 Newtons, per milliwatt(mW) of laser light.

to the attachment point. The tether itselfmay be a molecular motor, such as kinesinor myosin (22–24) (Figure 1a), a moleculeundergoing conformational transitions undertension, such as an RNA hairpin or protein(25–28) (Figure 1b), or the substrate to amotor, such as a DNA molecule packaged bya viral portal motor (Figure 1c) (20, 29).

How Do You Detect?

In each case, the action of the system—kinesinstepping along a microtubule, the RNA hair-pin unfolding, or the viral packaging motortranslocating DNA (Figure 1a-c)—is mon-itored from the motions of the bead in theoptical trap (or traps). Thus, the beads notonly serve as handles to a biological systembut also as probes by which its movementsare inferred. The sensitive detection of thebead position in its trap and, by extension,the calibrated measurement of displacementand force constitute a key feature of analyticaloptical tweezers. Several methods for measur-ing the position of a trapped particle exist andhave been covered extensively in the literature(18, 30). However, one technique, back-focalplane interferometry (31–33), has become thestandard in the field owing to its sensitivity,linearity, and speed in all three dimensions.

208 Moffitt et al.

Ann

u. R

ev. B

ioch

em. 2

008.

77:2

05-2

28. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Bol

ogna

on

06/2

7/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-10 ARI 28 April 2008 11:59

With this technique, it is in principle possi-ble to detect bead motions of less than 1 A onthe timescale of a millisecond or better, lim-ited only by the background electronic noiseof the detector (34).

How Do You Measure?

Optical tweezers data, in its raw form, typ-ically consist of various voltages from pho-todetectors and other components of theinstrument. Thus, before quantitative mea-surements can be made, these raw data mustbe converted to the actual physical param-eters of the system—the picoNewton forcesexerted on and the nanometer displacementsof the trapped bead. This calibration typi-cally involves measuring the response of thetrapped bead to a known force; however, thereare methods for measuring the optical forcedirectly (35). The standard calibration tech-niques are well reviewed in References 18and 30.

Unfortunately, even with calibrated dis-placements and forces, these values alone can-not determine the actual motion of the biolog-ical system. It is often necessary to also knowthe elastic behavior of the tethered moleculeand measure both its extension—the end-to-end distance from one attachment point to theother—and the tension—inferred from theoptical force exerted on the bead by the trap(36, 37). An alternative technique that provesuseful when the compliance of the tether or itsextension or attachment point are not knownis force feedback—a mode of operation in whichthe position of the optical trap relative to thesecond attachment point is adjusted to main-tain a constant optical force on the bead atall times (38). In this case, the bead displace-ment within the trap remains fixed, and it isthe displacement of the trap relative to thesecond attachment point that directly mea-sures the change in extension of the moleculeowing to the action of the biological system.In essence, a trap operated under force feed-back has zero stiffness because motions of theattachment point relative to the trap do not

result in changes in the applied force. Recentwork has shown that it is possible to achievethe same effect optically by exploiting nonlin-ear regions of the trapping potential (39).

The ability to detect movements of thebiological system, as inferred from the mo-tions of the bead handles or the motion ofthe optical trap, defines the spatial resolutionof the instrument, and the timescale of de-tectable motions defines its temporal resolu-tion. As is discussed below, these two conceptsare inextricably linked. In the following sec-tion, we review recent technical advances thathave pushed the resolution of optical tweez-ers to the angstrom spatial scale on the sec-ond timescale. For readers interested in fur-ther discussion of optical tweezers and use ofthe standard assay for biological studies, seeReferences 18–21, 26, 40–46.

IMPROVEMENTSIN RESOLUTION

Recently, major advances have been made inthe development of ultrastable and low-noiseoptical tweezers. These developments havecome not only in the form of better instru-mentation and technique but also in a betterunderstanding of the fundamental and phys-ical limitations to the sensitivity of opticaltweezers. In this section, we review these ex-citing advances.

Why Do You Want BetterResolution?

Many fundamental processes in the cell aremechanical in nature and occur by discretephysical movements, for example, the stepsof molecular motors along cytoskeletal fila-ments, the incorporation of one nucleotide toa nascent nucleic acid chain in transcription orreplication, or the unfolding and degradationof a protein domain by the proteasome. Of-ten, the size of these displacements is dictatedby the inherent periodicity of the substrate onwhich these systems act—the 8-nm repeat ofmicrotubules or the 3.4-A distance between

www.annualreviews.org • Recent Advances in Optical Tweezers 209

Ann

u. R

ev. B

ioch

em. 2

008.

77:2

05-2

28. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Bol

ogna

on

06/2

7/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-10 ARI 28 April 2008 11:59

adjacent base pairs in double-stranded DNA(dsDNA). More generally, however, theseprocesses can be viewed as reactions along anenergy landscape in which the reaction co-ordinate corresponds to a physical distance(17, 47, 48). The discrete motion in these pro-cesses stems from the fact that the states alongthese reaction pathways are highly localizedminima in this energy landscape. As a result,the direct observation and measurement ofthese discrete displacements, and deviationsfrom this discreteness, can reveal importantdetails on the energy landscape underlyingthese processes.

Although the size of such displacements iscertainly of fundamental interest, the abilityto detect discrete steps can also reveal detailedinformation on the governing kinetics. Whendiscrete steps are obscured by experimentalnoise, one can at most measure the averagerates of these processes and in some cases thefluctuations in these rates (48–50). By con-trast, the direct detection of discrete stepsallows one to measure the individual timeintervals between these events—the dwelltimes—and in turn compile their full prob-ability distribution. The dwell time distribu-tion contains far more information than thatavailable in the average rate, providing a sta-tistical measure of the kinetics (48, 50). In thecase of molecular motors, for instance, suchtechniques can be used to estimate the numberof kinetic transitions in their mechanochem-ical cycles (48, 49); for motors with multi-ple subunits, they can also reveal the detailedcoordination between the different subunits(51). With direct observation of the size andkinetics of the steps of a molecular motor, itis possible to reconstruct the detailed mecha-nism of the motor.

Unfortunately, these discrete movementshave only been detectable by traditional op-tical trap measurements in limited instances.For cytoskeletal motors such as kinesin ormyosin, the detection of steps is facilitated bythe relatively large step sizes taken by thesemotors and by the experimental geometry inwhich the motor is directly tethered to the

bead in the optical trap (Figure 1a). In this sit-uation, it is not uncommon to resolve steps ofa few nanometers with dwells as short as onlya few tens of milliseconds, resolution that ismore than sufficient for the detailed study ofthese motors (22–24). By contrast, resolvingthe motions of molecular motors that translo-cate along nucleic acid is considerably moredifficult, as these motors most likely movein steps on the single-base pair scale—a dis-tance of only 3.4 A on dsDNA. Moreover,some nucleic acid motors, such as the bac-terial DNA translocase FtsK (52, 53), can bequite fast, moving with speeds as high as 3–5 kb/s and requiring high temporal resolution.Regrettably, the experimental geometries thatuse nucleic acid as a tether to the trappedbead (Figure 1b,c) also suffer from worse spa-tial and temporal resolution, as described be-low. Thus, the detailed study of nucleic acidmotors has required significant improvementin resolution over traditional techniques. Be-yond their application to molecular motors,techniques with very high spatial and tempo-ral resolution will be necessary to study themany protein conformational changes at thecore of essential allosteric processes and en-zymatic functions. Although protein confor-mational changes as big as several nanome-ters are rare, changes in the subnanometerrange are ubiquitous. Thus, the ability to de-tect motions on the angstrom scale on a widerange of timescales, milliseconds to seconds,will allow the direct observation and study ofthe physical motions that underlie conforma-tional changes in a wide class of proteins.

What Limits Resolution?

The spatial resolution of optical tweezers islimited by drift and noise from a variety ofsources, which can be categorized as either ex-perimental, i.e., all noise stemming from theenvironment or from components of the in-strument, or Brownian, noise stemming fromthe fundamental thermal forces that gener-ate the fluctuations of the trapped objectsthemselves. Experimental noise is typically

210 Moffitt et al.

Ann

u. R

ev. B

ioch

em. 2

008.

77:2

05-2

28. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Bol

ogna

on

06/2

7/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-10 ARI 28 April 2008 11:59

caused by drift and fluctuations in the sam-ple stage or micropipette (Figure 1), result-ing in motions of the trapped bead, whichare indistinguishable from the motions gen-erated by the biological system. Alternatively,motions of the optical trap itself or fluctu-ations in the laser power can lead to simi-lar false signals. All of these problems canbe addressed through better instrumentation(18, 34).

Thermal forces, by contrast, cannot beavoided and provide a fundamental limit tothe resolution of a given experiment. To un-derstand the origin of this limit, consider theforces acting on the optically trapped bead. Atevery moment in time, the bead positions it-self in the optical trap to balance three forces:the force from the optical trap, the tensionon the biological system, and the spontaneousand random thermal force from the aqueousenvironment surrounding the trapped bead.2

To monitor the response of the biological sys-tem, the tension on the tether must be known,but this is never measured directly, only in-ferred from the measured optical force. Thus,this Brownian force introduces an uncertaintyin the applied tension on the biological sys-tem, which fundamentally limits the resolu-tion of an optical tweezers experiment. Fortu-nately, the effect of the Brownian force can becalculated, and experimental parameters canbe tuned to reduce its effect as we discuss be-low. In practice, experimental noise sourcesmust first be addressed before the instrumentcan attain a resolution limited only by funda-mental Brownian noise.

How Do You RemoveExperimental Noise?

The recent advances in high-resolution opti-cal tweezers suggest that experimental noise

2Viscous and inertial forces are typically negligible on thelong timescales associated with the dynamics of most bi-ological systems. For example, the inertial and viscoustimescales for a 1-μm polystyrene bead, in a trap of stiffness0.1 pN/nm, are ∼100 ns and ∼100 μs, respectively (54).

Detection laser: anadditional focusedlaser beam used fordetecting motion ofa bead but not fortrapping

mainly results from environmental factors.Temperature drift, mechanical and acousticvibrations in the room, as well as backgroundelectronic noise, all can couple into the in-strument and affect resolution (34). Criteriafor selecting a quiet environment have beendiscussed in References 18 and 34. In additionto improving the instrument environment,techniques for addressing any remainingexperimental noise have proven necessary toenhance the resolution of optical tweezers.In general, two approaches have been used:monitoring and subtracting this noise frommeasurements (55, 56) or designing theinstrument such that it is isolated from or in-sensitive to these sources of noise (34, 36, 57,58).

In experimental layouts in which the sur-face of a sample chamber is the second attach-ment point (Figure 1a), Nugent-Glandorf &Perkins (55) have shown that it is possible tomonitor and correct for the drift and fluc-tuations in the environment by introducinga second detection laser (30) to monitor theposition of the chamber surface. A fiducialmark is either deposited (55) or microfabri-cated (56) onto the chamber surface, allow-ing the detection laser to sensitively measurethe position of this mark. Drift and fluctu-ations can then be removed by either sub-tracting the motion of this fiducial mark fromthe data or through active feedback of thestage position. Carter et al. (56) have recentlydemonstrated that this approach can stabilizea stage to better than 1 A in three dimen-sions with a feedback bandwidth of 100 Hz.Approaches such as this should also be ableto stabilize systems that use micropipettes(Figure 1b).

The second approach is to isolate theinstrument completely from the samplechamber by the use of two optical traps, as inFigure 1c (34, 36, 57, 58). Because the systemof study is levitated above the chamber sur-face, slow drift in that surface has no effect onthe measurement. Furthermore, by formingthese dual traps from the same laser beam,the instrument can be effectively decoupled

www.annualreviews.org • Recent Advances in Optical Tweezers 211

Ann

u. R

ev. B

ioch

em. 2

008.

77:2

05-2

28. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Bol

ogna

on

06/2

7/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-10 ARI 28 April 2008 11:59

SNR: signal-to-noise ratio

from any drift in beam position. Remarkably,fluctuations in the two beam positions can stilloccur owing to slight changes in the index ofrefraction of the air from small density fluctu-ations and turbulence (34, 36, 58), in a similarfashion to the twinkling of stars at night. Thisatmospheric noise can be controlled either byhousing the crucial optics in an atmospherewith a smaller index of refraction for whichfluctuations produce smaller deviations in thebeam direction, such as helium (34, 58), or ina partial vacuum, or by reducing the differ-ential length of the optical paths traveled bythe two lasers (34, 36). In this latter case, anyatmospheric fluctuations in the common op-tical path will result in the same deviations forboth beams, thus producing no change in thedistance between the traps and no artifactualsignals.

How Do You AddressBrownian Noise?

As mentioned above, the thermal force thatengenders Brownian motion is the secondsource of resolution-limiting noise and pro-vides a fundamental limit to the resolution ofan experiment, even in the absence of exper-imental noise. Although thermal fluctuationscan never be completely eliminated, their de-pendence on experimentally tunable param-eters, such as bead size and tether stiffness,allows their effect on resolution to be min-imized. To this end, it is useful to under-stand how the resolution of an optical tweez-ers measurement depends on the choice ofexperimental parameters. A commonly usedfactor for assessing resolution is the signal-to-noise ratio (SNR), a dimensionless ratio ofthe size of a displacement signal to the noisethat may obscure it (Figure 2a). Although itsderivation falls outside the scope of this re-view and has been extensively treated in theliterature (36, 37), the expression for the SNRfor an instrument with one optical trap, inmeasurements carried out on sufficiently slow

timescales,3 is illuminating:

SNR ≤ κtether��√

4kB TBγ, 1.

where κtether is the stiffness of the biologicaltether, i.e., the slope of the force-extensioncurve of the tether at a given tension, �� is thesize of the physical displacement of the biolog-ical system, T is the temperature of the aque-ous bath, kB is the Boltzmann constant, B isthe bandwidth of the measurement—half therate at which data are collected, and γ = 6πηris the drag coefficient of the trapped bead. η

is the viscosity of the medium, and r is theradius of the trapped bead. In the perfect in-strument with no experimental noise, the ob-served SNR will be equal to the right-handside of this equation.

As Equation 1 shows, increasing the tetherstiffness, κtether , increases the SNR becausestiffer molecules are better transducers of me-chanical movement, whereas decreasing thetemperature, the drag coefficient of the bead,or the measurement bandwidth increases theresolution because reducing these factors de-creases the Brownian noise. Despite its ther-mal origin, decreasing the temperature doeslittle to lower the thermal noise because thebiologically permissive temperature range is asmall percentage of the absolute temperature.Decreasing bead size improves resolution be-cause smaller beads fluctuate faster than largerbeads and thus contribute less Brownian noiseon the measurement bandwidth (36, 37).4 Un-fortunately, for technical reasons, it can bedifficult to use beads with diameters smaller

3This equation is valid for motions slower than the viscoustimescale mentioned above. This timescale, set by the ratioof the drag coefficient to the trap stiffness (τ = γ /κ), pro-vides a fundamental limit to the temporal resolution of anoptical tweezers because the trapped bead acts a low-passfilter, averaging faster motions.4Although the total amount of noise over all frequenciesis fixed by the equipartition theorem (36, 37), small beadsdistribute this noise over a wider frequency range, leavingless noise at the low frequencies in which measurementsare made.

212 Moffitt et al.

Ann

u. R

ev. B

ioch

em. 2

008.

77:2

05-2

28. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Bol

ogna

on

06/2

7/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-10 ARI 28 April 2008 11:59

Figure 2Spatial resolution versus temporal resolution.(a) The signal-to-noise ratio (SNR) for a given stepis defined as the ratio of the mean distance betweensteps, �x, to the noise in a given step

√〈x2〉. The

information in the step clearly depends not juston the SNR of the data, both sets of steps picturedhere have a SNR of 4, but also on the number ofuncorrelated points that make up the step, 100 forthe red circles and 3 for the blue squares. (b) Fora given set of experimental parameters, Equation 2defines a region of step size and step duration thatare observable. Steps that are larger and slowerthan the solid blue line, for example, could beobserved with a two-trap system with equally sized1-μm beads and a 1-μm long dsDNA moleculeheld at a tension of 10 pN, assuming that a SNRof 4 and that on average 3 uncorrelated pointsper step are required for observation. The solidblack lines correspond to the possible step sizesand dwell times (using the fact that �� = v〈τ 〉)that would produce the lowest observed velocity,v, for Escherichia coli RNA polymerase (RNAP)(1 bp/s) (58), the bacteriophage ϕ29 (10 bp/s)(133), or the DNA translocase FtsK (300 bp/s)(52). The circles correspond to the observed(58) or estimated step sizes of these motors(52, 133). Thus, although 1-bp steps could beobserved for RNAP, the lowest velocity observedfor FtsK would require the steps to be at least∼6 bp in size for the same experimental setup.−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−→

than several hundred nanometers (34).5

Finally, a smaller measurement bandwidthimproves resolution because it allows addi-tional averaging of the fluctuations in the sig-nal, although the lowest value of the band-width is limited by the speed of the enzyme,as discussed below. Interestingly, there is nodependence in Equation 1 on trap stiffness,despite the fact that beads in stiffer traps fluc-tuate less. The reason is that beads in stiffertraps are also less sensitive to the motions ofthe biological system. Remarkably, these twofactors exactly cancel when taking the SNR,leading to the finding that trap stiffness has

5The strength of the optical trap, stiffness per power,and the sensitivity of back-focal-plane interferometry bothreach a maximum with a bead whose diameter is roughlyequal to the wavelength of the trapping light in water anddrop off quickly for larger or smaller beads (34).

N = 100

N = 3

RNAP

FtsK

Φ29Not observable

Observable

Ste

p s

ize

(bp

)D

ista

nce

(ar

bit

rary

un

its)

Time (arbitrary units)

Average dwell time (s)

SNR = Δx (√<x2>)–1a

b

Δx

<x2>

0

2

4

6

8

10

12

0.01 0.1 1

no effect on the Brownian limit to resolution.Because optical tweezers operating in forcefeedback can be essentially thought of as op-tical traps with zero stiffness, force feedbackdoes not offer any improvement in this reso-lution limit.

Are Two Traps Better than One?

An expression similar to the one above hasbeen derived by Moffitt et al. (36) for thetwo-trap geometry in Figure 1c. It wouldappear at first glance that the addition of a

www.annualreviews.org • Recent Advances in Optical Tweezers 213

Ann

u. R

ev. B

ioch

em. 2

008.

77:2

05-2

28. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Bol

ogna

on

06/2

7/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-10 ARI 28 April 2008 11:59

second bead subject to Brownian fluctuationswould increase the total noise in the system.This is true, it turns out, only when the beadsare treated independently of each other. Inreality, the dynamics of the two beads are cor-related because they are tethered together.Recent theoretical work and experimentalverification by the authors (36) demonstratethat these correlations can be exploited tominimize the effect of Brownian noise on thespatial resolution of dual-trap systems. Thisimprovement in resolution stems from thefact that fluctuations can be divided into twocomponents: (a) symmetric, where the beadsmove in the same direction, i.e., in phase,maintaining a constant separation, or (b) an-tisymmetric, where they move in oppositedirections, i.e., out of phase, stretching orcompressing the molecule tethering them.Because the displacements associated with theactivity of a biological system alter the ex-tension of the tether, only the antisymmet-ric component of this noise contributes tolimiting resolution. Thus, by monitoring themotion of both beads simultaneously, it is pos-sible to discard the noise caused by symmetricmotion with a differential measurement, im-proving the resolution of the system.

When the motion of both beads is mon-itored in this differential fashion, the SNRfor a dual-trap system takes an identical formto that of Equation 1, but with γ replaced byγeff = γ1γ2/(γ1 + γ2), where γ1,2 are the dragcoefficients of each of the beads (36). Becauseγeff is smaller than either γ1,2, the spatial reso-lution on a given bandwidth is higher for a sys-tem with two optical traps in which the move-ments of the beads in both traps are monitoredthan in a system with a single optical trap.Thus, a dual-trap geometry in which themotion of both beads is monitored is the bestexperimental geometry for limiting the effectsof thermal noise (36). Physically, the reasonfor this improvement is that both beads areinvolved in the dissipation of thermal energyand thus dissipate this energy more quickly,producing less fluctuation on the measure-ment bandwidth just as a smaller bead.

What ResolutionHas Been Achieved?

The recent advances in instrument stabil-ity and measurement, detailed above, havepushed the spatial resolution of optical tweez-ers to the angstrom scale. With two beads intwo optical traps tethered by a single piece ofdsDNA as in Figure 1c, stepping movementsof only 3.4 A at 0.5 to 1 s per step can nowbe detected (34, 36, 58). With this resolutionAbbondanzieri et al. (58) were able to mea-sure the single base pair motions of Escherichiacoli RNA polymerase (RNAP) when slowed to∼1 nucleotide/s with low nucleotide concen-tration under assisting forces as low as 18 pN.In these experiments, one of the two traps wasmuch stiffer than the other so that the motionsof the bead in the stiffer trap could be ignored.Moffitt et al. (36) demonstrated that by form-ing two optical traps of comparable stiffnessand by monitoring the motion of both beads,as discussed above, it is possible to obtain sim-ilar resolution even when using larger beadsand when applying several times lower tensionon the DNA tether.

What Is Needed to Observe Steps?

The measurement bandwidth, B, is a conve-nient unit from an instrumentation perspec-tive because it characterizes the rate at whichdata are collected and averaged; however, forthe purpose of studying periodic biologicalprocesses, such as the stepping of a molec-ular motor, the average period of the system(the time per step of a molecular motor, forexample) 〈τ 〉 is a much more intuitive quan-tity. The bandwidth of the measurement canbe related to this quantity by 2B = N/〈τ 〉,where N is the number of uncorrelated mea-surements per average dwell.6 N ≥ 1 sets thelimit on the measurement bandwidth if onedoes not wish to average over the individual

6The factor of two originates in the subtle differencebetween the bandwidth of the measurement and the fre-quency at which the data are sampled, namely, fsamp = 2B.

214 Moffitt et al.

Ann

u. R

ev. B

ioch

em. 2

008.

77:2

05-2

28. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Bol

ogna

on

06/2

7/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-10 ARI 28 April 2008 11:59

periods of the system. Using this expressionit is possible to show that the observation of astep of size �� with an average dwell time of〈τ 〉 requires that

��√

〈τ 〉 ≥√

2kB TNγ

κtetherSNR. 2.

This relationship makes explicit the inverserelationship between spatial and temporal res-olution. Observation of single base pair stepsat 1 step per second is just as difficult as ob-serving steps of two base pairs at 0.25 s per stepor steps of 10 bp at 10 ms per step. Figure 2b

further illustrates this point.Although Equation 2 provides a more in-

tuitive connection to the physical motions ofthe biological system, it does not determinewhat SNR or what value of N are required forthe observation of steps and the extraction ofunbiased kinetic information from these steps.Recent empirical work was aimed at address-ing these questions. For example, one recentstudy (59) has suggested that a SNR of 4 orbetter is required for the measurement of stepsizes given the most commonly used algo-rithm, the pairwise distribution. From otherwork, a value of N ≈ 1 appears to be sufficientfor the measurement of steps (34, 36, 58), al-though little is reported on how this affectsthe determination of the dwell times of steps.Similar results have been obtained for someof the more sophisticated step-finding algo-rithms (60–63). The empirical approaches ofthese studies provide useful rules of thumb fordesigning experiments, but a systematic the-oretical study of the affect of SNR and N onstep size and dwell time measurements is stillneeded.

What about Accuracy?

An underappreciated effort that has been oc-curring in parallel to the development of high-resolution trapping techniques has been thecontinued improvements in the accuracy oftheir calibration. The degree of accuracy—the relative error between the measured valueand actual value—needed in a given experi-

Pairwisedistribution: ahistogram of thedistances betweenevery two points inthe trajectory of amolecular motor

ment depends on the particular application.For instance, a 10% calibration error might betolerated in the measurement of RNAP steps;a measurement of a 1.1-bp step size would notbe interpreted as inconsistent with a 1-bp step.In contrast, FtsK has been estimated to step inincrements of 12 bp (52); direct observationsof these steps, with ∼10% calibration error re-vealing an 11-bp step size, would impact howthe data are interpreted and how models of themotor operation are constructed. More im-portantly, although it is widely believed thatnucleic acid translocases move in discrete in-crements that are integer multiples of a sin-gle base pair, this need not be true. High-resolution optical tweezers have the potentialto test this long-standing paradigm throughdirect observations, but only if the accuracyof these measurements is high. Thus, if mod-els are to be constructed faithfully and withoutbias, efforts to improve the accuracy of opticaltweezers measurements must go hand in handwith efforts to improve resolution. Recent ad-vances in this direction have focused on de-veloping new calibration techniques (64) andreducing systematic errors in existing calibra-tion procedures (54, 65, 66). Though the na-ture of these developments is too technical towarrant detailed discussion in this review, theyare an important contribution to the field,and the interested reader is encouraged toread the literature describing these advances(54, 64).

HYBRID AND NOVELINSTRUMENTS

In parallel to the advances in high-resolutionoptical trapping, the field has also seen thedevelopment of novel and hybrid opticaltweezers. These instruments significantly ex-pand the capabilities of the traditional opticaltweezers, allowing the manipulation of morecomplicated biological systems, the control ofadditional variables such as torque, or the si-multaneous use of additional forms of single-molecule manipulation and detection. In thissection, we review these recent advances.

www.annualreviews.org • Recent Advances in Optical Tweezers 215

Ann

u. R

ev. B

ioch

em. 2

008.

77:2

05-2

28. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Bol

ogna

on

06/2

7/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-10 ARI 28 April 2008 11:59

Time sharing: aprocess that rapidlyswitches a singlelaser betweenmultiple locations toform multiple opticaltraps

Acousto-opticdeflector (AOD): acrystal that acts as atunable diffractiongrating under theapplication ofradiofrequencysound waves

Why Create Hybridand Novel Instruments?

The motion of many molecular motors isinherently three-dimensional, involving notonly translocation along a molecular track, butmovements in orthogonal directions and evenrotation around this track. Moreover, thismotion may involve many simultaneous con-formational changes within the protein. Tra-ditional optical tweezers, which typicallyproject all motion onto a single axis, are notgood tools for the investigation of these addi-tional motions. Furthermore, inside the cell,molecular motors do not work in isolationbut are part of large complexes that involvethe tightly coordinated interaction of multiplemolecules at multiple locations. Again, tradi-tional optical tweezers, which can manipulateonly a limited number of molecules with oneor two optical traps, are poorly suited to recre-ate this aspect of the cellular environment.Both of these limitations may be overcomeby introducing new manipulation and mea-surement capabilities into the standard opticaltweezers assay.

Why Manipulate Morethan One Molecule?

To illustrate one method in which novel formsof manipulation have been integrated into op-tical tweezers, consider the technical prob-lems associated with using optical tweezers tostudy the condensation of the bacterial chro-mosome. The bacterial chromosome, and allof its associated condensing proteins, is a com-plex, dynamic structure whose integrity anddegree of condensation involves interactionsbetween many distal regions of DNA and theparticipation of many proteins. Among theseproteins is the nucleoid structuring proteinH-NS, a dimeric protein that has two in-dependent DNA-binding domains (67). Theorientation and mode of binding of thesedomains—whether they bind adjacently to thesame DNA segment or bridge distal segments,forming loops—and thus their role in the con-

densation of the bacterial chromosome werenot known (67). To complicate matters, ini-tial single-molecule studies using only a sin-gle piece of DNA revealed that H-NS did notcompact a single DNA molecule but ratherextended it (68, 69).

To address this problem, Dame et al. (70)constructed an optical tweezers with four in-dependently controllable optical traps, whichthey term the Quad-trap (Figure 3a). Withthis geometry it was possible for the au-thors to individually trap four polystyrenebeads and use them to manipulate two in-dependent DNA molecules. With this abil-ity, the authors were able to determine thatH-NS can only bridge DNA strands whenthey are overlapping before incubation withH-NS but not when they are coated with H-NS and then overlapped. This observationsuggests that a single H-NS dimer bridgesand condenses distal regions of the chromo-some by DNA-protein-DNA interactions asin Figure 3a. These authors were also ableto measure the force needed to both unzipand shear a DNA-H-NS-DNA assembly, andbecause they designed their instrument tomaintain the high spatial and temporal res-olution of traditional optical tweezers, theyobserved the discrete steps in which the con-densed DNA unzipped. The use of four opti-cal traps allowed these experiments to be con-ducted in a straightforward and simple fash-ion not possible with only one or two opticaltraps.

How Do You Create NovelOptical Potentials?

The four traps in the Quad-trap instrumentused by Dame et al. (70) are formed by split-ting a single laser via polarization, using onepolarization to form one of the optical traps,and then time sharing the other polariza-tion with an acousto-optic deflector (AOD) toform the other three traps (70) (Figure 3b).Time sharing a single laser between multipleoptical traps is possible as long as the laser isreturned to the same location faster than the

216 Moffitt et al.

Ann

u. R

ev. B

ioch

em. 2

008.

77:2

05-2

28. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Bol

ogna

on

06/2

7/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-10 ARI 28 April 2008 11:59

H-NSa

RF AOD

PBS

PBS PBS

PD

OBJ CND

b

F

Figure 3Control of multimolecular complexes with multiplexed optical tweezers. (a) The formation of four opticaltraps allows control of a DNA-H-NS-DNA complex not possible with only a single piece of DNA. Bymeasuring force with one of the optical traps (red ), it is possible to follow the unzipping of this complexon the nanometer scale. (b) The formation of four optical traps is accomplished with an acousto-opticdeflector (AOD). When driven by different radio frequencies (RFs), this device rapidly shares one of thepolarizations of the laser (blue), split with a polarizing beam splitter (PBS), between three different traps.The other polarization (red ) is not shared, allowing light from this trap to be easily separated withanother PBS after the traps are formed with a high-numerical objective (OBJ), and the scattered light iscollected with a condenser (CND). Projecting this light onto a position sensitive photodector (PD)allows precise measurement of the response of the bead in the trap that is not time shared. Note that thisis a schematic diagram only, necessary relay lenses have been left out for clarity (70).

time it takes for the bead to diffuse away fromthat location, typically on the order of tens ofmilliseconds. For simplicity, Dame et al. de-tect the force only in the single trap that is nottime shared (Figure 3b), although Guilfordet al. (71) have shown it is possible to performthe complicated signal processing necessaryto extract the force and position of each of thebeads in multiplexed optical traps.

Time-sharing techniques using AODs andother methods have been described previously(30, 71–73) and have been used to generatemultiple optical traps for the study of nonpro-cessive motors such as myosin (74). Further-more, it is possible to use this technique togenerate novel optical potentials in additionto multiple optical traps. If the physical lo-cation of the time-shared traps are within thediameter of the trapped bead, and if the trap isshifted fast enough, the bead observes an av-erage two-dimensional potential that can bearbitrarily shaped given the average durationand light intensity of the trap at each loca-tion. Such novel optical potentials have been

Spatial lightmodulator: thisdevice allows precise,dynamic control overthe phase andintensity of differentregions of anincident laser

used to form regions of constant force for thestudy of the dynamics of DNA relaxation (75),and also so-called keyhole traps (76), poten-tials that are capable of trapping a bead boundto a cytoskeletal filament while also aligningthe filament within the trap. These potentialshave been useful for the study of the forcesinvolved in the dynamics of microtubule andactin filament assembly (60, 76, 77).

Another powerful approach to create mul-tiple optical traps or novel optical potentialsis to use spatial light modulators to imprint aphase image of the desired traps in the trap-ping light, which is then transformed by theobjective into the desired location and num-ber of optical traps.7 Such holographic opti-cal tweezers offer enormous versatility in thenumber, size, shape, and three-dimensionalposition of these traps and have thus inspired

7Lenses also work by modifying the phase of different re-gions of the incoming light; thus, spatial light modulatorscan be thought of as computer-controlled lenses of arbi-trary shape.

www.annualreviews.org • Recent Advances in Optical Tweezers 217

Ann

u. R

ev. B

ioch

em. 2

008.

77:2

05-2

28. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Bol

ogna

on

06/2

7/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-10 ARI 28 April 2008 11:59

a large body of research and development(for brief reviews see References 78 and 79).Unfortunately, there is currently no methodfor detecting the force and response on thetrapped beads that rivals the resolution andbandwidth offered by back-focal plane in-terferometry or comparable laser detectionmethods. The problem is inherent in the useof phase to encode the individual optical traps.Because the phase of the light is also changedupon scattering from the bead, it is difficult todiscriminate and separate the scattered lightfrom each of the individual trapped parti-cles. One possible solution is to follow Dameet al. (70) by introducing an additional op-tical trap formed from a unique polarizationor wavelength to measure force and use theholographic optical tweezers for manipulationonly. Alternatively, the additional laser neednot be powerful enough to form an opticaltrap but could simply be a detection laser (30).Such a design would provide greater versatil-ity, allowing the force and position of any ofthe optically trapped beads to be monitored.

Why Add Torqueto an Optical Tweezer?

Not all hybrid instruments that involve newforms of manipulation have focused on theassembly and manipulation of complex bio-logical systems. For example, much effort hasbeen dedicated to extending optical tweezersto the measurement of additional degrees offreedom, such as rotation. These new instru-ments offer the potential for studying bio-logical motors that generate twist and torquein conjunction with force. In addition to themore obvious examples, such as the flagellarmotor (80), F1F0 ATP synthase (81, 82), andtype II topoisomerases (83–85), many canon-ically linear motors are also believed to pro-duce rotation. For example, motors such asDNAPs and RNAPs, which involve the syn-thesis of a nucleic acid chain from a DNAtemplate, most likely follow the helical pitchof this template, generating a relative rota-tion of one turn for every 10.5 bp translo-

cated. Moreover, the mechanisms of multi-meric, ring-shaped nucleid acid motors mayalso involve rotation. In the chromosome seg-regation DNA translocase FtsK (86), or the vi-ral DNA-packaging motor of bacteriophageϕ29 (87), the symmetry mismatch betweenthe multimeric ring of the motor and thatof the DNA helix suggests that motor andDNA twist relative to each other to main-tain specific contacts during translocation.Thus, measurements of the twist per distancetranslocated can provide insight into the num-ber of subunits within the motor, the types ofcontacts these subunits make with the track,and even the spatial order in which the sub-units make these contacts during transloca-tion. Given these considerations, it is of greatinterest to find methods to control twist andmeasure torque while simultaneously measur-ing linear motion and force with the resolu-tion of traditional optical tweezers.

How Do You Add Rotational Controlto Optical Tweezers?

Bryant et al. (88) introduced one of the firstmethods for providing controlled rotationinto an optical tweezers. By adding a rotat-ing micropipette (Figure 4a), the authorswere able to induce twist on a molecule (88)or molecules of DNA (89) while simultane-ously measuring force and extension of themolecule. Although this method of addingtwist is relatively simple and need not involveextensive modification to an existing instru-ment, it does not provide a direct measure-ment of torque. The authors solved this prob-lem by adding a small “rotor” bead to theside of the DNA. The engineering of a nickon one strand of the DNA near this rotorbead allowed the single bonds of the oppositestrand to act as molecular swivels. It turns outthat the angular velocity of this bead—drivenby the torque on the DNA and balancedby the torque generated by viscous drag—is directly proportional to the torque storedon the DNA. By over- or undertwisting theDNA and then watching the rate at which the

218 Moffitt et al.

Ann

u. R

ev. B

ioch

em. 2

008.

77:2

05-2

28. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Bol

ogna

on

06/2

7/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-10 ARI 28 April 2008 11:59

a b c

χe

L R

N S

τF

NS

ττL R

Figure 4Methods for generating torque in optical tweezers. (a) Rotation of a micropipette allows the addition oftwist to a single molecule of DNA and the application of a torque, τ , which is measured by the rotation ofa small “rotor” bead attached to the DNA. (b) Optically trapped paramagnetic beads develop a magneticdipole in the presence of a magnetic field, which can be used to rotate the bead by rotating the magneticfield. (c) Optically trapped birefrigent particles, such as quartz microcylinders, change the polarization ofthe trapping light, generating an angular momentum change, and, thus, a torque if the extraordinary axis(χe ) of the particle is rotated with respect to the incident polarization. If linearly polarized trapping lightis used (with no net angular momentum owing to the equal parts left (L) and right (R) of circularlypolarized light), the optical torque can be measured directly by determining the amount of right and leftcircularly polarized light in the forward-scattered light (102, 103).

rotor bead dissipated this torque, the authorswere able to determine the constant of pro-portionality between the twist applied to themolecule and the torque stored—the torsionalrigidity of the molecule (88). Moreover, theresearchers were able to use this experimentalsetup to probe the detailed phase diagram ofdsDNA, determining the forces and torquesat which it undergoes transitions between dif-ferent structural forms (88). By introducing asecond DNA molecule and using the rotatingmicropipette to braid these molecules, Stoneet al. (89) were able to also use a rotating mi-cropipette to study the mechanism of type IItopoisomerases.

Instead of introducing a rotating mi-cropipette, which may not be practical in alloptical setups, it is also possible to trap a para-magnetic bead and use external magnets torotate this bead in the optical trap (90–92)(Figure 4b). The addition of magnets to astandard optical tweezers can be relativelysimple and provides a straightforward methodfor adding the ability to twist biomoleculesto a preexisting instrument. Unfortunately,there is again no direct method for measur-ing the torque on the paramagnetic bead. Asbefore, it is possible to introduce a small rotor

bead to measure the torque on DNA; how-ever, because the magnetic bead is free to ro-tate in the optical trap, it is also possible totrack its rotation, assuming the bead is visiblyasymmetric (91). By monitoring the phase de-lay between the rotation angle of this defectand that of a rotating magnetic field caused bythe viscous drag on the rotating bead, it is pos-sible to calibrate the torque for each bead (91).The potential drawbacks of such a method arethat it must be repeated for each trapped beadand requires sensitive measure of the relativeangle between the bead defect and the mag-netic field. Recently, Crut et al. (93) used suchmagneto-optical tweezers to measure the re-laxation of plectonemic DNA.

A more instrumentally challenging, butmore powerful, method for introducingtorque-generating capabilities into opticaltweezers is to use the angular momentum car-ried by light. Torque may be generated onthe trapped particle either by transfer of thisangular momentum to the particle via par-tial absorption of the trapping light (94–96)or by addition of angular momentum intothe trapping light by the trapping particle it-self (97, 98). In the latter case, the asymme-try of the particle, either because of its shape

www.annualreviews.org • Recent Advances in Optical Tweezers 219

Ann

u. R

ev. B

ioch

em. 2

008.

77:2

05-2

28. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Bol

ogna

on

06/2

7/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-10 ARI 28 April 2008 11:59

Birefringence: anoptical asymmetry inwhich light ofdifferent polarizationtravels at differentspeeds through amaterial

Extraordinary axis:a birefringent crystalaxis that allows lightto travel fastest whenits polarization isaligned with this axis

(99–101) or its birefringent properties (97, 98,102), changes the angular momentum of theincoming beam, producing a torque on theparticle to conserve angular momentum. Thebenefit of such a method is that it does notrequire the absorption of trapping light, al-lowing the use of relatively transparent parti-cles, which in turn reduces the effect of laserheating (97). Furthermore, if the angular mo-mentum is transferred into the polarizationof the light, as opposed to the spatial dis-tribution of the light, then it is possible tomeasure exactly the torque on the particle viachanges in the polarization state of the scat-tered light (102, 103). Specifically, if trappingis accomplished with linearly polarized light,which carries no net angular momentum, thetorque on the particle can be measured bydetermining the excess of right-circularly toleft-circularly polarized light after its inter-action with the trapped particle (Figure 4c).In direct analogy to force and position, op-tical measurement of torque and rotation ofthe linear polarization of light can be donewith very high bandwidth, making it possibleto create torque feedback systems that rotatethe polarization of the trapping light dynami-cally to keep the applied torque on the systemconstant (102). For a comprehensive reviewon optical torque see Reference 104.

One potential drawback of such opticaltorque wrenches is the need for uniform bire-fringent particles that have the extraordinaryaxis oriented correctly with the desired axisof the experiment and with any asymme-tries within the trapping particle (98). Re-cently, Deufel et al. (105) have demonstrateda powerful solution to this problem by fab-ricating micron-sized quartz cylinders withthe extraordinary axis of the quartz uniformlyaligned along the short axis of the cylinder.The cylindrical shape ensures that the longaxis will align in the axial direction of the op-tical trap and properly orient the extraordi-nary axis of the cylinder with the trappingpolarization (105). Furthermore, the authorsamino-functionalize only one surface of thecylinder, controlling the attachment point of

the biological system and further ensuringthat rotation occurs around the correct axis.The power of their experimental setup wasdemonstrated by measuring simultaneouslythe torque on and the extension of a singlepiece of dsDNA while twisting the moleculeunder a constant optical force. In this sim-ple proof-of-principle experiment (105), sev-eral of the mechanical features of dsDNAwere measured simultaneously, such as thetwist stretch coupling (106, 107), the torsionalrigidity (88), and the critical torque for thephase transition between B-form DNA andsupercoiled DNA (88).

Finally, it is also possible to introduce twistor apply torque to a biological system with afixed axis by simply moving an optical trapin a circle around the fixed axis of the sam-ple. Pilizota et al. (108) have constructed anoptical tweezers with this capability for thestudy of the bacterial flagellar motor. Becausethe radius of the circle ascribed by the motionof the optical tweezers is known, it should bepossible to calculate the applied torque fromthe applied optical force. For systems in whichthere is no simple fixed axis, one may use a sec-ond optical trap to form this rotation point(109).

What Other forms of ManipulationCan Be Integrated withOptical Tweezers?

In the above examples, it is the bead in theoptical trap that is subjected to some addi-tional form of manipulation; however, therehave been several recent examples of employ-ing optical tweezers to measure the responseof other manipulation techniques. For exam-ple, Keyser et al. (110, 111) recently integratedoptical tweezers with nanopores and demon-strated that it is possible to pull on a singlepiece of DNA partially threaded through ananopore and driven by an applied electricfield, a feat recently repeated by Trepagnieret al. (112). By measuring the required force asa function of the voltage across the nanopore,Keyser et al. were able to measure the

220 Moffitt et al.

Ann

u. R

ev. B

ioch

em. 2

008.

77:2

05-2

28. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Bol

ogna

on

06/2

7/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-10 ARI 28 April 2008 11:59

effective free charge per base pair of DNA(111). In addition, Huisstede et al. (113) inte-grated a glass micropipette scanning probe ina single-trap optical tweezers. They demon-strated that by correlating the response of thebead in the optical trap to the position ofa scanning probe coated with antibodies todigoxigenin, it was possible to localize indi-vidual digoxigenin modifications on a singlepiece of DNA.

How is Single-MoleculeFluorescence Used?

The field of fluorescence detection is basedon the ability to optically excite small fluo-rophores, small organic molecules or proteins,or semiconductors nanocrystals, i.e., quantumdots, which emit photons of a longer wave-length when they relax back to their groundstate. One of the major benefits of this tech-nique is that the system is largely noninvasive:Typical fluorophores are small, and the exci-tation and detection are completely optical,allowing far-field excitation and imaging. Fur-thermore, there are a wide range of methodsfor illuminating the sample: epi-illuminationin which the entire sample plane is bathedin the excitation light, confocal illuminationin which a small diffraction-limited spot isexcited, or total internal reflection where anevanescent wave is used to excite a small re-gion near the surface of the sample slide.Useful single-molecule measurements can bemade with large numbers of dyes, throughmultiply labeled molecules, or through onlya few dyes. In the case of small numbers ofdyes, photophysical interactions between thedyes such as Forster resonant energy trans-fer (FRET) or self-quenching can be used toprobe small distance changes or can serve asmolecular on-off switches. Additionally, withadvances in single-fluorophore imaging, it ispossible to track single dyes on the nanometerand millisecond scales. For reviews of the ap-plication of single-molecule fluorescence seeReferences 21, 45, 114, and 115.

FRET: Forsterresonant energytransfer

How Do You Use Fluorescenceto Visualize Molecules inOptical Tweezers?

Fluorescence as a method for visualizing sin-gle molecules has been utilized since thevery first experiments that manipulated DNA(116). In the first optical tweezers assay, a longpiece of DNA (tens of microns) was labeledwith ethidium bromide and stretched betweentwo beads held in two optical traps (117). Therelaxation of the DNA as one bead was re-leased could be followed by simply watch-ing the movements of the fluorescently la-beled DNA. Similar assays have been used tonot only study the polymer physics of DNA(118, 119), but have also been adapted to thestudy of molecular motors, such as the heli-case complex RecBCD (120–122). Controlledfluid flow was used to extend the DNA, andchanges in its length visualized by fluores-cence microscopy revealed the progress of themotor. Although such assays are powerful, theoptical trap is used only as a means of ma-nipulating the DNA and holding it stationaryagainst a fluid flow.

In one recent example, which illustratesthe possibilities of combining fluorescenceimaging of multiply labeled molecules withforce measuring optical tweezers, the elasticproperties of a single piece of dsDNA hetero-geneously coated with filaments of Rad51, animportant component in eukaryotic homolo-gous recombination, were investigated (123).In this assay, fluorescently labeled Rad51 fila-ments could be imaged on a piece of dsDNAtethered between two optically trapped beads(Figure 5a). The response of both the fluo-rescent filaments and the bare DNA to an ap-plied load could be followed simultaneouslywith fluorescence microscopy and the deflec-tion of the trapped beads, allowing the authorsto dissect the elastic properties of each com-ponent from a measurement of the heteroge-neous system (123). This example illustratesone application of combined optical tweezers-fluorescence measurements: the ability to de-marcate different parts of a larger molecular

www.annualreviews.org • Recent Advances in Optical Tweezers 221

Ann

u. R

ev. B

ioch

em. 2

008.

77:2

05-2

28. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Bol

ogna

on

06/2

7/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-10 ARI 28 April 2008 11:59

FADP

ADP

a

b

Figure 5Fluorescence in optical tweezers. (a) Multiple fluorophores (stars) can beused to tag different regions of a heterogeneous biological tether allowingthe individual response of each section of the tether to optical forces to beprobed with video microscopy. (b) Single fluorescently labeled moleculescan be used to correlate the bound nucleotide state and the force-generating transition of a molecular motor such as myosin. Total internalreflection illumination limits the spatial extent of the excitation light(blue), lowering the fluorescent background from free nucleotides (124).

complex and measure the response of each toan applied load.

Can You Use Single Fluorophoreswith Optical Tweezers?

Although the use of multiple fluorophoresto label single molecules offers many advan-tages, there are many applications of fluores-cence imaging that require the use of only oneor two dyes. The earliest example of single-molecule fluorescence combined with opti-cal tweezers is still perhaps one of the mostimpressive accomplishments in the field. Ishi-jima et al. (124) suspended an actin fiber be-tween two optical traps and measured thedeflection of the fiber by individual myosinheads. However, the true tour de force of thisexperiment was in the incorporation of totalinternal reflection fluorescence excitation andsingle-fluorophore detection into this system,

which allowed the investigators to simultane-ously detect single fluorescently labeled ATP(ATP-Cy3) bind to the myosin (124). This ex-periment made it possible to directly corre-late the nucleotide binding state of the mo-tor and its power stroke (Figure 5b). Theauthors confirmed that binding of nucleotideoccurred simultaneously with the disengage-ment of myosin from actin and that, rathercontroversially (125), the power stroke of themotor did not correspond to the release ofADP but was a different and subsequent ki-netic transition (124).

The widespread use of single-moleculefluorescence measurements and opticaltweezers force and displacement measure-ments has been slowed, however, owingto several technical problems. The maindifficulty is that the lifetime of the fluores-cent dyes decreases dramatically because ofthe absorption of a near-infrared trappingphoton by the fluorescent excited state(126). This effect was not experienced inthe study just described, as the rigidity ofactin allowed the investigators to use longfilaments and thus separate the region inwhich single fluorophores were imaged andthe locations of the optical traps by ∼10 μm.Unfortunately, the detection of molecularmotions is greatly inhibited by more flexiblepolymers, such as dsDNA, of this lengthbecause the stiffness of the tether decreaseswith increased length and lower stiffnessdegrades resolution, as explained above.Nevertheless, it is possible to still measureand apply force with the optical trap on longpieces of DNA and then use fluorescencechanges owing to FRET, for example,as a sensitive measure of distance (126a).In applications requiring close proximitybetween the optical trap and the fluorescentdyes, this effect on lifetime can be mitigatedsomewhat by introducing antioxidants (126),judicious choice of dye (the fluorophore Cy3is more sensitive to this effect than TMRor Alexa555, for example) (127, 128), oreven using orthogonal polarizations for the

222 Moffitt et al.

Ann

u. R

ev. B

ioch

em. 2

008.

77:2

05-2

28. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Bol

ogna

on

06/2

7/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-10 ARI 28 April 2008 11:59

trapping light and the fluorescence excitation(126). An alternative method is to avoidsimultaneous exposure of the fluorophores toboth light sources simultaneously. Brau et al.(129) have developed a method in which theyinterlace the excitation light and the trappinglight, with no deleterious effects on trappingor fluorescence as long as the modulationrate is high enough. They have accomplisheda 20-fold improvement in the lifetime ofCy3 molecules and have demonstrated theprinciple of their method by simultaneouslydetecting conformational changes in DNAhairpins under load using single fluorophores(129) and FRET (130).

CONCLUSIONS ANDPERSPECTIVES

As described in this review, optical trappingtechniques have rapidly advanced in recentyears along two fronts: the improvement ofspatial resolution and accuracy, and the hy-bridization of techniques for the simultaneousmeasurement of multiple observables. Theformer developments have been motivatedby the desire to understand conformationalchanges in proteins and nucleic acids at theirmost fundamental spatial and temporal scales.High-resolution optical tweezers can now di-rectly detect these motions, together with theactual stepping of a number of molecular mo-tors and the conformational changes result-ing from the mechanical unfolding of proteinsand nucleic acids. More importantly, the im-proved spatial and temporal resolution makeit now possible to study a much larger class ofbiological processes that occur at these lengthscales, a regime traditionally accessible onlyto high-resolution structural methods. For ex-ample, conformational changes that lie at theheart of many cellular processes, such as en-zyme catalysis, cell signaling, and ion channelgating, all involve motions of protein domainsat the angstrom scale and therefore internalgeneration of force (17). The ability to resolveslow motions on this spatial scale should thus

allow the direct study of these motions andthe effect of force on their kinetics. This de-velopment thus provides the exciting prospectof complementing the static pictures providedby atomic structures with kinetic measure-ments of conformational changes on similarlength scales.

In spite of these recent advances inresolution, there is still much room forimprovement. All high-resolution instrumen-tation studies report that experimental noisebecomes limiting at longer timescales, typi-cally one second or longer (34, 36, 58). Al-though the source of this noise remains insome cases unclear, its elimination will beessential to achieve even higher spatial res-olution on those slow time regimes. Theextension of angstrom resolution to fastertimescales is certainly possible as well, asEquations 1 and 2 demonstrate. However, thisnext leap in resolution will most likely re-quire more radical ideas and perhaps novelexperimental geometries to allow manipula-tion of shorter and stiffer tethers and smallerbeads.

The second trend in the advancement ofoptical tweezers has been to extend the tech-nique to the study of increasingly complex sys-tems to better duplicate their in vivo physiol-ogy. The development of hybrid techniqueshas aimed to address the challenge of prob-ing these complex macromolecular assem-blies. Although the measurement of displace-ment and force along one axis afforded bytraditional optical tweezers has proven insuf-ficient to fully characterize the dynamics ofsome systems, hybrid methods provide theopportunity for additional readouts such asforce and displacement along complemen-tary axes, torque, and angle, or fluorescenceposition and orientation. These capabilitiesshould allow one to measure the generationof force and torque simultaneously on manymolecular motors. Moreover, not only willit be possible to correlate nucleotide statewith motor translocation, as Ishijima et al.(124) did with myosin and single-fluorophore

www.annualreviews.org • Recent Advances in Optical Tweezers 223

Ann

u. R

ev. B

ioch

em. 2

008.

77:2

05-2

28. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Bol

ogna

on

06/2

7/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-10 ARI 28 April 2008 11:59

detection of labeled nucleotides, but it mayalso be feasible to use optical techniques tocontrol the nucleotide state as well, withcaged-ATP molecules, for instance, or moreexotic photoswitchable molecules (131, 132).Finally, in the future we expect hybrid tech-niques to unleash the full arsenal of single-fluorophore imaging methodologies (115).

Many of the techniques described in thisreview have still only had proof-of-principledemonstrations, in part, because they are sonew, but also because the instrumentationinvolved is relatively complex. Nevertheless,these techniques hold great promise to ad-dress many fundamental problems in biol-ogy at the single-molecule level. The nextstep in the development of optical tweez-

ers instruments will no doubt combine thesehybrid techniques with the highest spatialand temporal resolution. Although most likelyvery difficult to construct, these instrumentspromise an unprecedented amount of infor-mation. Imagine following the discrete stepsof a molecular motor at high resolution whilecorrelating these steps with domain motionsrevealed by single-molecule fluorescence. Al-most 35 years ago Arthur Ashkin was experi-menting with intense light and latex beads sus-pended in water, a work that spawned this field(1). Today, as the field matures and impor-tant technical advances continue to be made,it is clear that optical tweezers will increas-ingly become an important weapon in the fullarsenal of the biochemist.

SUMMARY POINTS

1. Optical tweezers have seen several exciting technological advances in the past years,which will allow researchers to dissect the molecular mechanism of more complicatedbiological systems in greater detail.

2. By addressing both experimenal and Brownian sources of noise, it is now possible toresolve angstrom scale motions on the second timescale.

3. The use of multiple optical traps or novel optical potentials is allowing researchers tomanipulate biological systems of greater complexity.

4. Several methods have been demonstrated for introducing the capability of applyingand measuring torque simultaneously with force in optical tweezers.

5. Many of the initial hurdles in combining optical tweezers and single-molecule fluo-rescence have been overcome, and these hybrid instruments promise to provide anunprecedented level of detail into the inner workings of molecular motors.

FUTURE ISSUES

1. Further improvement in the resolution of optical tweezers is possible but will requiregreater attention to experimental sources of noise and the use of smaller beads andshorter tethers.

2. Because of the relatively uninvasive nature of optical tweezers, it should be possibleto integrate this technique with other forms of single-molecule manipulation andmeasurement.

3. The full power of the myriad hybrid techniques now awaits application to a widerange of biological systems.

224 Moffitt et al.

Ann

u. R

ev. B

ioch

em. 2

008.

77:2

05-2

28. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Bol

ogna

on

06/2

7/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-10 ARI 28 April 2008 11:59

DISCLOSURE STATEMENT

The authors are not aware of any biases that might be perceived as affecting the objectivity ofthis review.

ACKNOWLEDGMENTS

We thank M. Nollmann, C.L. Hetherington, and W. Cheng for a critical reading of themanuscript. J.R.M. acknowledges the National Science Foundation’s Graduate Research Fel-lowship and Y.R.C. the Burroughs Welcome Fund’s Career Awards at the Scientific Interfacefor funding.

LITERATURE CITED

1. Ashkin A. 1970. Phys. Rev. Lett. 24:156–9

Firstdemonstration ofstable opticaltrapping ofmicron-sizeddielectric objects inthree dimensionsusing a single-beamgradient opticaltrap.

2. Ashkin A, Dziedzic JM, Bjorkholm JE, Chu S. 1986. Opt. Lett. 11:288–903. MacDonald MP, Spalding GC, Dholakia K. 2003. Nature 426:421–244. Pauzauskie PJ, Radenovic A, Trepagnier E, Shroff H, Yang P, Liphardt J. 2006. Nat.

Mater. 5:97–1015. Agarwal R, Ladavac K, Roichman Y, Yu GH, Lieber CM, Grier DG. 2005. Opt. Express

13:8906–126. Furst EM. 2005. Curr. Opin. Colloid Interface Sci. 10:79–867. Meiners JC, Quake SR. 1999. Phys. Rev. Lett. 82:2211–148. Henderson S, Mitchell S, Bartlett P. 2001. Phys. Rev. E 64:0614039. Henderson S, Mitchell S, Bartlett P. 2002. Phys. Rev. Lett. 88:088302

10. Hough LA, Ou-Yang HD. 2002. Phys. Rev. E 65:02190611. Liphardt J, Dumont S, Smith SB, Tinoco I Jr, Bustamante C. 2002. Science 296:1832–3512. Wang GM, Sevick EM, Mittag E, Searles DJ, Evans DJ. 2002. Phys. Rev. Lett. 89:05060113. Trepagnier EH, Jarzynski C, Ritort F, Crooks GE, Bustamante CJ, Liphardt J. 2004.

Proc. Natl. Acad. Sci. USA 101:15038–4114. Collin D, Ritort F, Jarzynski C, Smith SB, Tinoco I, Bustamante C. 2005. Nature

437:231–3415. Bustamante C. 2005. Q. Rev. Biophys. 38:291–30116. Bustamante C, Liphardt J, Ritort F. 2005. Phys. Today 58:43–48

A detailed reviewof the myriadmechanicalprocesses inbiochemistry.

17. Bustamante C, Chemla YR, Forde NR, Izhaky D. 2004. Annu. Rev. Biochem.

73:705–48

A comprehensivereview of thetheory of opticaltrapping and thepractical concernsassociated withconstructing anoptical tweezers.

18. Neuman KC, Block SM. 2004. Rev. Sci. Instrum. 75:2787–80919. Bustamante C, Macosko JC, Wuite GJ. 2000. Nat. Rev. Mol. Cell Biol. 1:130–3620. Bustamante C, Bryant Z, Smith SB. 2003. Nature 421:423–2721. Greenleaf WJ, Woodside MT, Block SM. 2007. Annu. Rev. Biophys. Biomol. Struct.

36:171–9022. Svoboda K, Schmidt CF, Schnapp BJ, Block SM. 1993. Nature 365:721–2723. Finer JT, Simmons RM, Spudich JA. 1994. Nature 368:113–1924. Mehta AD, Rief M, Spudich JA, Smith DA, Simmons RM. 1999. Science 283:1689–9525. Liphardt J, Onoa B, Smith SB, Tinoco I Jr, Bustamante C. 2001. Science 292:733–3726. Zhuang X. 2005. Annu. Rev. Biophys. Biomol. Struct. 34:399–41427. Tinoco I, Li PTX, Bustamante C. 2006. Q. Rev. Biophys. 39:325–6028. Cecconi C, Shank EA, Bustamante C, Marqusee S. 2005. Science 309:2057–60

www.annualreviews.org • Recent Advances in Optical Tweezers 225

Ann

u. R

ev. B

ioch

em. 2

008.

77:2

05-2

28. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Bol

ogna

on

06/2

7/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-10 ARI 28 April 2008 11:59

29. Smith DE, Tans SJ, Smith SB, Grimes S, Anderson DL, Bustamante C. 2001. Nature413:748–52

30. Visscher K, Gross SP, Block SM. 1996. IEEE J. Sel. Top. Quantum Electron. 2:1066–7631. Allersma MW, Gittes F, deCastro MJ, Stewart RJ, Schmidt CF. 1998. Biophys. J.

74:1074–8532. Gittes F, Schmidt CF. 1998. Opt. Lett. 23:7–933. Pralle A, Prummer M, Florin EL, Stelzer EHK, Horber JKH. 1999. Microsc. Res. Tech.

44:378–86

A detaileddiscussion of howto build ahigh-resolutiondual trap opticaltweezers.

34. Bustamante C, Chemla YR, Moffitt JR. 2008. In Single-Molecule Techniques: A

Laboratory Manual, ed. PR Selvin, T Ha, pp. 297–324. Cold Spring Harbor, NY:Cold Spring Harb. Lab.

35. Smith SB, Cui Y, Bustamante C. 2003. Methods Enzymol. 361:134–62

Detailedcalculation of thethermal resolutionlimit for systemswith one and twotraps.

36. Moffitt JR, Chemla YR, Izhaky D, Bustamante C. 2006. Proc. Natl. Acad. Sci. USA

103:9006–1137. Gittes F, Schmidt CF. 1998. Eur. Biophys. J. 27:75–8138. Visscher K, Block SM. 1998. Methods Enzymol. 298:460–8939. Greenleaf WJ, Woodside MT, Abbondanzieri EA, Block SM. 2005. Phys. Rev. Lett.

95:20810240. Svoboda K, Block SM. 1994. Annu. Rev. Biophys. Biomol. Struct. 23:247–8541. Ashkin A. 1997. Proc. Natl. Acad. Sci. USA 94:4853–6042. Mehta AD, Rief M, Spudich JA. 1999. J. Biol. Chem. 274:14517–2043. Hormeno S, Arias-Gonzalez JR. 2006. Biol. Cell 98:679–9544. Ritort F. 2006. J. Phys. Condens. Matter 18:R531–8345. Cornish PV, Ha T. 2007. ACS Chem. Biol. 2:53–6146. Neuman KC, Lionnet T, Allemand JF. 2007. Annu. Rev. Mater. Res. 37:33–6747. Bustamante C, Keller D, Oster G. 2001. Acc. Chem. Res. 34:412–2048. Kolomeisky AB, Fisher ME. 2007. Annu. Rev. Phys. Chem. 58:675–9549. Schnitzer MJ, Block SM. 1995. Cold Spring Harb. Symp. Quant. Biol. 60:793–80250. Shaevitz JW, Block SM, Schnitzer MJ. 2005. Biophys. J. 89:2277–8551. Asbury CL, Fehr AN, Block SM. 2003. Science 302:2130–3452. Saleh OA, Perals C, Barre FX, Allemand JF. 2004. EMBO J. 23:2430–3953. Pease PJ, Levy O, Cost GJ, Gore J, Ptacin JL, et al. 2005. Science 307:586–9054. Berg-Sorensen K, Flyvbjerg H. 2004. Rev. Sci. Instrum. 75:594–61255. Nugent-Glandorf L, Perkins TT. 2004. Opt. Lett. 29:2611–1356. Carter AR, King GM, Ulrich TA, Halsey W, Alchenberger D, Perkins TT. 2007. Appl.

Opt. 46:421–2757. Shaevitz JW, Abbondanzieri EA, Landick R, Block SM. 2003. Nature 426:684–8758. Abbondanzieri EA, Greenleaf WJ, Shaevitz JW, Landick R, Block SM. 2005. Nature

438:460–6559. Wallin AE, Salmi A, Tuma R. 2007. Biophys. J. 93:795–80560. Kerssemakers JWJ, Munteanu EL, Laan L, Noetzel TL, Janson ME, Dogterom M.

2006. Nature 442:709–1261. Milescu LS, Yildiz A, Selvin PR, Sachs F. 2006. Biophys. J. 91:1156–6862. Milescu LS, Yildiz A, Selvin PR, Sachs F. 2006. Biophys. J. 91:3135–5063. Carter BC, Vershinin M, Gross SP. 2008. Biophys. J. 94:306–1964. Tolic-Norrelykke SF, Schaffer E, Howard J, Pavone FS, Julicher F, Flyvbjerg H. 2006.

Rev. Sci. Instrum. 77:10310165. Berg-Sorensen K, Oddershede L, Florin EL, Flyvbjerg H. 2003. J. Appl. Phys. 93:3167–

76

226 Moffitt et al.

Ann

u. R

ev. B

ioch

em. 2

008.

77:2

05-2

28. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Bol

ogna

on

06/2

7/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-10 ARI 28 April 2008 11:59

66. Berg-Sorensen K, Peterman EJG, Weber T, Schmidt CF, Flyvbjerg H. 2006. Rev. Sci.Instrum. 77:063106

67. Dame RT. 2005. Mol. Microbiol. 56:858–7068. Amit R, Oppenheim AB, Stavans J. 2003. Biophys. J. 84:2467–7369. Dame RT, Wuite GJL. 2003. Biophys. J. 85:4146–4870. Dame RT, Noom MC, Wuite GJL. 2006. Nature 444:387–9071. Guilford WH, Tournas JA, Dascalu D, Watson DS. 2004. Anal. Biochem. 326:153–6672. K. Visscher GJBJJK. 1993. Cytometry 14:105–1473. Molloy JE. 1998. Methods Cell Biol. 55:205–1674. Molloy JE, Burns JE, Sparrow JC, Tregear RT, Kendrick-Jones J, White DC. 1995.

Biophys. J. 68:S298–30575. Nambiar R, Gajraj A, Meiners JC. 2004. Biophys. J. 87:1972–8076. Kerssemakers JWJ, Janson ME, van der Horst A, Dogterom M. 2003. Appl. Phys. Lett.

83:4441–4377. Footer MJ, Kerssemakers JWJ, Theriot JA, Dogterom M. 2007. Proc. Natl. Acad. Sci.

USA 104:2181–8678. Grier DG. 2003. Nature 424:810–1679. Grier DG, Roichman Y. 2006. Appl. Opt. 45:880–8780. Berg HC. 2003. Annu. Rev. Biochem. 72:19–5481. Nakamoto RK, Ketchum CJ, Al-Shawi MK. 1999. Annu. Rev. Biophys. Biomol. Struct.

28:205–3482. Yoshida M, Muneyuki E, Hisabori T. 2001. Nat. Rev. Mol. Cell Biol. 2:669–7783. Corbett KD, Berger JM. 2004. Annu. Rev. Biophys. Biomol. Struct. 33:95–11884. Charvin G, Strick TR, Bensimon D, Croquette V. 2005. Annu. Rev. Biophys. Biomol.

Struct. 34:201–1985. Nollmann M, Crisona NJ, Arimondo PB. 2007. Biochimie 89:490–9986. Massey TH, Mercogliano CP, Yates J, Sherratt DJ, Lowe J. 2006. Mol. Cell 23:457–6987. Simpson AA, Tao YZ, Leiman PG, Badasso MO, He YN, et al. 2000. Nature 408:745–5088. Bryant Z, Stone MD, Gore J, Smith SB, Cozzarelli NR, Bustamante C. 2003. Nature

424:338–4189. Stone MD, Bryant Z, Crisona NJ, Smith SB, Vologodskii A, et al. 2003. Proc. Natl. Acad.

Sci. USA 100:8654–5990. Sacconi L, Romano G, Ballerini R, Capitanio M, De Pas M, et al. 2001. Opt. Lett.

26:1359–6191. Romano G, Sacconi L, Capitanio M, Pavone FS. 2003. Opt. Commun. 215:323–3192. Claudet C, Bednar J. 2005. Appl. Opt. 44:3454–5793. Crut A, Koster DA, Seidel R, Wiggins CH, Dekker NH. 2007. Proc. Natl. Acad. Sci.

USA 104:11957–6294. He H, Friese MEJ, Heckenberg NR, Rubinsztein-Dunlop H. 1995. Phys. Rev. Lett.

75:826–2995. Friese MEJ, Enger J, Rubinsztein-Dunlop H, Heckenberg NR. 1996. Phys. Rev. A

54:1593–9696. Friese MEJ, Nieminen TA, Heckenberg NR, Rubinsztein-Dunlop H. 1998. Opt. Lett.

23:1–397. Friese MEJ, Nieminen TA, Heckenberg NR, Rubinsztein-Dunlop H. 1998. Nature

394:348–5098. Singer W, Nieminen TA, Gibson UJ, Heckenberg NR, Rubinsztein-Dunlop H. 2006.

Phys. Rev. E 73:021911

www.annualreviews.org • Recent Advances in Optical Tweezers 227

Ann

u. R

ev. B

ioch

em. 2

008.

77:2

05-2

28. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Bol

ogna

on

06/2

7/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-10 ARI 28 April 2008 11:59

99. Galajda P, Ormos P. 2001. Appl. Phys. Lett. 78:249–51100. Bishop AI, Nieminen TA, Heckenberg NR, Rubinsztein-Dunlop H. 2003. Phys. Rev. A

68:033802101. Oroszi L, Galajda P, Kirei H, Bottka S, Ormos P. 2006. Phys. Rev. Lett. 97:058301102. La Porta A, Wang MD. 2004. Phys. Rev. Lett. 92:190801103. Nieminen TA, Heckenberg NR, Rubinsztein-Dunlop H. 2001. J. Mod. Opt. 48:405–13104. Parkin S, Knoner G, Singer W, Nieminen TA, Heckenberg NR, et al. 2007. In Methods

in Cell Biology, ed. MW Berns, KO Greulich, pp. 525–61. New York: Academic105. Deufel C, Forth S, Simmons CR, Dejgosha S, Wang MD. 2007. Nat. Methods 4:223–25106. Gore J, Bryant Z, Nollmann M, Le MU, Cozzarelli NR, Bustamante C. 2006. Nature

442:836–39107. Lionnet T, Joubaud S, Lavery R, Bensimon D, Croquette V. 2006. Phys. Rev. Lett.

96:178102108. Pilizota T, Bilyard T, Bai F, Futai M, Hosokawa H, Berry RM. 2007. Biophys. J. 93:264–

75109. Bingelyte V, Leach J, Courtial J, Padgett MJ. 2003. Appl. Phys. Lett. 82:829–31110. Keyser UF, van der Does J, Dekker C, Dekker NH. 2006. Rev. Sci. Instrum. 77:105105111. Keyser UF, Koeleman BN, van Dorp S, Krapf D, Smeets RMM, et al. 2006. Nat. Phys.

2:473–77112. Trepagnier EH, Radenovic A, Sivak D, Geissler P, Liphardt J. 2007. Nano Lett. 7:2824–

30113. Huisstede JHG, Subramaniam V, Bennink ML. 2007. Microsc. Res. Tech. 70:26–33114. Weiss S. 1999. Science 283:1676–83115. Toprak E, Selvin PR. 2007. Annu. Rev. Biophys. Biomol. Struct. 36:349–69116. Smith SB, Finzi L, Bustamante C. 1992. Science 258:1122–26117. Chu S. 1991. Science 253:861–66118. Perkins TT, Quake SR, Smith DE, Chu S. 1994. Science 264:822–26119. Perkins TT, Smith DE, Chu S. 1994. Science 264:819–22120. Handa N, Bianco PR, Baskin RJ, Kowalczykowski SC. 2005. Mol. Cell 17:745–50121. Spies M, Bianco PR, Dillingham MS, Handa N, Baskin RJ, Kowalczykowski SC. 2003.

Cell 114:647–54122. Bianco PR, Brewer LR, Corzett M, Balhorn R, Yeh Y, et al. 2001. Nature 409:374–78123. Mameren J, Modesti M, Kanaar R, Wyman C, Wuite GJ, Peterman EJ. 2006. Biophys.

J. 91:L78–80124. Ishijima A, Kojima H, Funatsu T, Tokunaga M, Higuchi H, et al. 1998. Cell 92:161–71125. Wallace MI, Molloy JE, Trentham DR. 2003. J. Biol. 2:4126. Van Dijk MA, Kapitein LC, Van Mameren J, Schmidt CF, Peterman EJG. 2004. J. Phys.

Chem. B. 108:6479–84126a. Hohng S, Zhou R, Nahas MK, Yu J, Schulten K, et al. 2007. Science 318:279–83127. Lang MJ, Fordyce PM, Engh AM, Neuman KC, Block SM. 2004. Nat. Methods 1:133–39128. Lang MJ, Fordyce PM, Block SM. 2003. J. Biol. 2:6129. Brau RR, Tarsa PB, Ferrer JM, Lee P, Lang MJ. 2006. Biophys. J. 91:1069–77130. Tarsa PB, Brau RR, Barch M, Ferrer JM, Freyzon Y, et al. 2007. Angew. Chem. Int. Ed.

Engl. 46:1999–2001131. Banghart M, Borges K, Isacoff E, Trauner D, Kramer RH. 2004. Nat. Neurosci. 7:1381–

86132. Volgraf M, Gorostiza P, Numano R, Kramer RH, Isacoff EY, Trauner D. 2006. Nat.

Chem. Biol. 2:47–52133. Chemla YR, Aathavan K, Michaelis J, Grimes S, Jardine PJ, et al. 2005. Cell 122:683–92

228 Moffitt et al.

Ann

u. R

ev. B

ioch

em. 2

008.

77:2

05-2

28. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Bol

ogna

on

06/2

7/08

. For

per

sona

l use

onl

y.

AR345-FM ARI 7 May 2008 14:43

Annual Review ofBiochemistry

Volume 77, 2008Contents

Prefatory Chapters

Discovery of G Protein SignalingZvi Selinger � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �1

Moments of DiscoveryPaul Berg � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 14

Single-Molecule Theme

In singulo Biochemistry: When Less Is MoreCarlos Bustamante � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 45

Advances in Single-Molecule Fluorescence Methodsfor Molecular BiologyChirlmin Joo, Hamza Balci, Yuji Ishitsuka, Chittanon Buranachai,and Taekjip Ha � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 51

How RNA Unfolds and RefoldsPan T.X. Li, Jeffrey Vieregg, and Ignacio Tinoco, Jr. � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 77

Single-Molecule Studies of Protein FoldingAlessandro Borgia, Philip M. Williams, and Jane Clarke � � � � � � � � � � � � � � � � � � � � � � � � � � � � �101

Structure and Mechanics of Membrane ProteinsAndreas Engel and Hermann E. Gaub � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �127

Single-Molecule Studies of RNA Polymerase: Motoring AlongKristina M. Herbert, William J. Greenleaf, and Steven M. Block � � � � � � � � � � � � � � � � � � � �149

Translation at the Single-Molecule LevelR. Andrew Marshall, Colin Echeverría Aitken, Magdalena Dorywalska,and Joseph D. Puglisi � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �177

Recent Advances in Optical TweezersJeffrey R. Moffitt, Yann R. Chemla, Steven B. Smith, and Carlos Bustamante � � � � � �205

Recent Advances in Biochemistry

Mechanism of Eukaryotic Homologous RecombinationJoseph San Filippo, Patrick Sung, and Hannah Klein � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �229

v

Ann

u. R

ev. B

ioch

em. 2

008.

77:2

05-2

28. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Bol

ogna

on

06/2

7/08

. For

per

sona

l use

onl

y.

AR345-FM ARI 7 May 2008 14:43

Structural and Functional Relationships of the XPF/MUS81Family of ProteinsAlberto Ciccia, Neil McDonald, and Stephen C. West � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �259

Fat and Beyond: The Diverse Biology of PPARγ

Peter Tontonoz and Bruce M. Spiegelman � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �289

Eukaryotic DNA Ligases: Structural and Functional InsightsTom Ellenberger and Alan E. Tomkinson � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �313

Structure and Energetics of the Hydrogen-Bonded Backbonein Protein FoldingD. Wayne Bolen and George D. Rose � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �339

Macromolecular Modeling with RosettaRhiju Das and David Baker � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �363

Activity-Based Protein Profiling: From Enzyme Chemistryto Proteomic ChemistryBenjamin F. Cravatt, Aaron T. Wright, and John W. Kozarich � � � � � � � � � � � � � � � � � � � � � �383

Analyzing Protein Interaction Networks Using Structural InformationChristina Kiel, Pedro Beltrao, and Luis Serrano � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �415

Integrating Diverse Data for Structure Determinationof Macromolecular AssembliesFrank Alber, Friedrich Förster, Dmitry Korkin, Maya Topf, and Andrej Sali � � � � � � � �443

From the Determination of Complex Reaction Mechanismsto Systems BiologyJohn Ross � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �479

Biochemistry and Physiology of Mammalian SecretedPhospholipases A2

Gerard Lambeau and Michael H. Gelb � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �495

Glycosyltransferases: Structures, Functions, and MechanismsL.L. Lairson, B. Henrissat, G.J. Davies, and S.G. Withers � � � � � � � � � � � � � � � � � � � � � � � � � � �521

Structural Biology of the Tumor Suppressor p53Andreas C. Joerger and Alan R. Fersht � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �557

Toward a Biomechanical Understanding of Whole Bacterial CellsDylan M. Morris and Grant J. Jensen � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �583

How Does Synaptotagmin Trigger Neurotransmitter Release?Edwin R. Chapman � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �615

Protein Translocation Across the Bacterial Cytoplasmic MembraneArnold J.M. Driessen and Nico Nouwen � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �643

vi Contents

Ann

u. R

ev. B

ioch

em. 2

008.

77:2

05-2

28. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Bol

ogna

on

06/2

7/08

. For

per

sona

l use

onl

y.

AR345-FM ARI 7 May 2008 14:43

Maturation of Iron-Sulfur Proteins in Eukaryotes: Mechanisms,Connected Processes, and DiseasesRoland Lill and Ulrich Mühlenhoff � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �669

CFTR Function and Prospects for TherapyJohn R. Riordan � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �701

Aging and Survival: The Genetics of Life Span Extensionby Dietary RestrictionWilliam Mair and Andrew Dillin � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �727

Cellular Defenses against Superoxide and Hydrogen PeroxideJames A. Imlay � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �755

Toward a Control Theory Analysis of AgingMichael P. Murphy and Linda Partridge � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �777

Indexes

Cumulative Index of Contributing Authors, Volumes 73–77 � � � � � � � � � � � � � � � � � � � � � � � �799

Cumulative Index of Chapter Titles, Volumes 73–77 � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �803

Errata

An online log of corrections to Annual Review of Biochemistry articles may be foundat http://biochem.annualreviews.org/errata.shtml

Contents vii

Ann

u. R

ev. B

ioch

em. 2

008.

77:2

05-2

28. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Bol

ogna

on

06/2

7/08

. For

per

sona

l use

onl

y.