Amplified stretch of bottlebrush-coated DNA innanofluidic channelsCe Zhang1 Armando Hernandez-Garcia23 Kai Jiang1 Zongying Gong1

Durgarao Guttula1 Siow Yee Ng1 Piravi P Malar1 Jeroen A van Kan1 Liang Dai4

Patrick S Doyle45 Renko de Vries23 and Johan R C van der Maarel14

1Department of Physics National University of Singapore 2 Science Drive 3 117542 Singapore 2Laboratory ofPhysical Chemistry and Colloid Science Wageningen University 6703 HB Wageningen The Netherlands 3Foodand Biobased Research Wageningen University 6700 AA Wageningen The Netherlands 4BioSystems andMicromechanics (BioSym) IRG Singapore MIT Alliance for Research and Technology (SMART) 117576 Singaporeand 5Department of Chemical Engineering Massachusetts Institute of Technology Cambridge MA 02139 USA

Received April 29 2013 Revised August 5 2013 Accepted August 9 2013

ABSTRACT

The effect of a cationic-neutral diblock polypeptideon the conformation of single DNA moleculesconfined in rectangular nanochannels is investi-gated with fluorescence microscopy An enhancedstretch along the channel is observed withincreased binding of the cationic block of the poly-peptide to DNA A maximum stretch of 85 of thecontour length can be achieved inside a channelwith a cross-sectional diameter of 200 nm and at a2-fold excess of polypeptide with respect to DNAcharge With site-specific fluorescence labelling itis demonstrated that this maximum stretch is suffi-cient to map large-scale genomic organizationMonte Carlo computer simulation shows that theamplification of the stretch inside the nanochannelsis owing to an increase in bending rigidity and thick-ness of bottlebrush-coated DNA The persistencelengths and widths deduced from the nanochanneldata agree with what has been estimated from theanalysis of atomic force microscopy images of driedcomplexes on silica

INTRODUCTION

Advances in nanofabrication have made it possible tofabricate quasi one-dimensional devices with cross-sectional diameters on the order of tens to hundreds ofnanometers These nanochannels serve as a platform forstudying among others single DNA molecules (1ndash8)Furthermore confinement in a nanospace result in signifi-cant modification of certain biophysical phenomena suchas the knotting probability of circular DNA and the effectof macromolecular crowding on the conformation and

folding of DNA (9ndash12) Nanochannels are also importantfor single-molecule technologies for mapping of large-scale genomic organization including restriction enzymecutting nick labelling and denaturation mapping (13ndash18)Advantages of the nanochannel platform are that theDNA molecules are in an equilibrium conformationhigh throughput can be achieved by using parallel arraysof channels and integration with lab-on-chip devicesFuture devices might involve nonequilibrium transloca-tion of DNA molecules through a pore or by a sensor aprocess that can be facilitated by a nanochannel platform(19) The molecules of interest are usually visualized withfluorescence microscopyThe resolution of the genome mapping technologies

relies on the ability to stretch DNA to a length close toits contour length Stretching of single DNA molecules togt60 of the contour length can be achieved by confine-ment in channels with cross-sectional diameters below100 nm (31415) Alternatively DNA molecules can bestretched by working at low ionic strength of lt1mM(4ndash6) At such low ionic strength the persistence lengthof the duplex exceeds 100 nm and is significantly largerthan the physiological value of around 50 nm Followingthe latter strategy of amplifying the stretch by an increasein bending rigidity we here propose to apply a polymercoating to the double-stranded DNA molecule The ad-vantage of this approach is that wider channels can beused to achieve a similar stretch In particular the lessstringent channel diameter allows the use of nanofluidicdevices produced by soft lithography and made of elasto-mer such as polydimethylsiloxane (PDMS) Elastomerchips have some advantages including the possibility ofobtaining multiple replicas by casting on a master stamp(3) Furthermore a wider nanofluidic channel systemallows in situ changes in environmental solution condi-tions which is required for DNA binding assays Using

To whom correspondence should be addressed Tel 65 65164396 Fax 65 67776126 Email johanmaarelgmailcom

Published online 3 September 2013 Nucleic Acids Research 2013 Vol 41 No 20 e189doi101093nargkt783

The Author(s) 2013 Published by Oxford University PressThis is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (httpcreativecommonsorglicensesby-nc30) which permits non-commercial re-use distribution and reproduction in any medium provided the original work is properly cited For commercialre-use please contact journalspermissionsoupcom

a novel device based on a cross-channel configurationrecently we have shown that DNA molecules trappedinside wider channels (250 nm) remain accessible forbinding proteins andor ligands (20) The wider channelsalso allow loading of DNA molecules by electrophoresisRecently a cationic-neutral diblock polypeptide has been

developed that qualifies as a suitable candidate for coatingof DNA (21) The cationic DNA binding block is relativelyshort and consist of 12 lysine residues (K12) The K12 blockis coupled to four repeats of a 100-amino acid long poly-peptide (C4) (22) The C4 block has an amino acid compos-ition similar to that of collagen Furthermore it ishydrophilic net electroneutral and behaves as a flexiblepolymer in aqueous solution The entire C4K12 diblockpolypeptide is produced using recombinant DNA technol-ogy by large-scale expression with high yields in yeast Thediblock copolymers are monodisperse with a total molecularweight of 384 kDa The polypeptide concentration is con-veniently expressed as the NP ratio that is the number ofpositively charged amino groups on the lysine K12 bindingblock divided by the number of phosphate groups on theDNA As shown in previous work complexation of DNAwith C4K12 does not result in aggregates consisting ofmultiple DNA molecules (21) Instead single DNA mol-ecules are uniformly coated by a dense bottlebrush asillustrated in Figure 1 We have verified that DNA metab-olism is not inhibited by coating with the polypeptide (seelsquoMaterials and Methodsrsquo section) Two effects are expectedto enhance stretching of the DNAndashC4K12 complex innanochannels a larger cross-sectional diameter and alarger bending rigidity or persistence length These effectswill be gauged from nanofluidics experiments atomic forcemicroscopy and Monte Carlo simulation

MATERIALS AND METHODS

Sample preparation

C4K12 polypeptide (Mw=384 kDa) was expressed in theyeast Pichia Pastoris as an exocellular protein and purifiedby NH4SO4 precipitation of the supernatant (2122) Thepurity and absence of degradation of the polypeptideshave been verified with sodium dodecyl sulphate-poly-acrylamide gel electrophoresis and matrix-assisted laserdesorption ionization time-of-flight mass spectrometer(MALDI-TOF) To verify the effect of polypeptidecoating on DNA metabolism we have done a digestionassay (21) As shown by the image of the gel in Figure 2bare DNA is degraded by DNAaseI in lt1min In thesame condition polypeptide-coated DNA with NP=75 is degraded in about an hour DNA metabolismis hence slowed down but not inhibited by the polypeptidecoating T4-DNA was purchased from Nippon GeneTokyo and used without further purification -DNANbBbvCI nicking endonuclease and VentR (exo-) DNApolymerase were purchased from New England BiolabsIpswich MA USA YOYO-1 and ChromaTide AlexaFluor 546-14-dUTP were purchased from InvitrogenCarlsbad CA USA -DNA was sequence-specificallystained with Alexa Fluor 546 (15) T4-DNA was stainedwith YOYO-1 with an intercalation ratio of 100 bp per dye

molecule For nanofluidics samples were prepared bydialysing solutions of T4-DNA against 10mM TrisHCland 1mM EDTA in micro-dialysers The TrisHCl con-centration is 10mM Tris adjusted with HCl to pH 85 ie29mM TrisCl and 71mM Tris The ionic strength of thebuffer was calculated with the Davies equation forestimating the activity coefficients of the ions and a dis-sociation constant pK=808 for Tris The polypeptidewas dissolved in the same buffer Solutions of polypeptideand DNA were subsequently mixed in equal volumes andincubated for 24 h at 277K The final DNA concentrationis 3mgl No anti-photo bleaching agent was used Atomicforce microscopy was done with pUC18 plasmid(2686 bp) pUC18 was purchased and linearized withHindIII restriction endonuclease (New England BiolabsIpswich MA USA) Solutions of plasmid and polypep-tide were mixed in equal volumes to a final concentrationof 02mg of DNAl and incubated for 24 h at 277K

Chip fabrication

The nanofluidic devices were fabricated by replication inPDMS of patterned master stamps (523) The stampswere made in hydrogen silsesquioxane (HSQ) resist (DowCorning Midland MI USA) using a lithography processwith proton beam writing (2425) The60 5 100 5 200 5 250 5 and 300 5 nm heightsof the positive channel structures on the stamps were

h

Rs C

K

4

12

DNA

Figure 1 Illustration of a bottlebrush formed by binding of a diblockpolypeptide copolymer to DNA The binding block of the copolymercontains 12 cationic lysine residues

1 2 3 11 12 134 5 6 7 8 9 10 14 15 17 18 1916 20 21

Figure 2 Degradation of pMTL23 (37 kb) by DNAaseI The plasmidand DNAseI concentrations are 72 nM and 88Uml respectivelyLanes 1 12 and 21 DNA molar weight marker Lanes 2ndash11 reactionproduct of bare DNA after 1 5 8 15 25 35 45 60min of incubationrespectively Lanes 13ndash20 as lanes 2ndash11 but for C4K12-coated DNAwith NP=75

e189 Nucleic Acids Research 2013 Vol 41 No 20 PAGE 2 OF 8

measured with atomic force microscopy (Dimension 3000Veeco Woodbury NY USA) The stamps were replicatedin PDMS followed by curing with a curing agent (SylgardDow Corning) at 338K for 24h (26) After both substrateswere plasma oxidized (Harrick Ossining NY USA) thePDMS replicas were sealed with glass slides The widthsof the channels in the replicas were measured with atomicforce microscopy and the values agreed with those obtainedfrom the scanning electron microscopy images of the masterstamps The nanochannels have a length of 60mm and rect-angular cross-sections of100 60 250 100 200 200 250 200 250 250 250300 300 300 and 500 300nm2 A fresh chip wasmade for every experiment

To check for possible adhesion of polypeptide toPDMS we have prepared a flat PDMS chip The chipwas plasma oxidized and covered with glass slides Asquare 2 2mm2 window of the PDMS surface was leftuncovered The chip was exposed to a solution of poly-peptide (1 gl) for the typical duration of a nanofluidicexperiment (30min) The glass slides were removed thechip was N2-dried and the surface was imaged withatomic force microscopy We did not observe a differencein surface roughness or a transition in height level fromthe previously covered to the exposed PDMS surface areaof the chip Accordingly there is no appreciable adhesionof polypeptide to PDMS

Fluorescence imaging

The stained DNA molecules dispersed in the relevantsolution were loaded into one of the two reservoirs con-nected by the nanochannels The DNA molecules weresubsequently driven into the channels by electrophoresisFor this purpose two platinum electrodes were immersedin the reservoirs and connected to an electrophoresispower supply with a relatively low voltage in the range01ndash10V (Keithley Cleveland OH USA) Once theDNA molecules were localized inside the nanochannelsthe electric field was switched off and the molecules wereallowed to relax to their equilibrium state for at least 60 sThe stained DNA molecules were visualized with a NikonEclipse Ti inverted fluorescence microscope equipped witha 200 W metal halide lamp a filter set and a 100 oilimmersion objective A ultraviolet light illuminationshutter controlled the exposure time Images were col-lected with an electron multiplying charge coupleddevice camera (Andor iXon X3) and the extension ofthe DNA molecules inside the channels was measuredwith imageJ software (httprsbinfonihgovij)

DNA combing

Polystyrene (Mw=280 kDa) is dissolved in toluene(Fisher Scientific Pittsburgh PA USA) at a concentra-tion of 100 gl Glass cover slips were cleaned by sonic-ation in 70 ethanol for 30min The cover slips werespin-coated with the solution of polystyrene for 30 s at2000 rpm A 5-ml droplet was spotted onto the cover slipand sheared with the help of a pipette tip along thesurface The DNA molecules were visualized with fluores-cence microscopy (27)

Atomic force microscopy

All imaging experiments were carried out at room tempera-ture in air with a Dimension 3000 atomic force microscopeVeeco Woodbury NY USA Images were acquired in thetapping mode with silicon (Si) cantilevers (spring constantof 20ndash100Nm) and operated below their resonance fre-quency (typically 230ndash410kHz) The images were flattenedand the contrast and brightness were adjusted for optimumviewing conditions A 20-ml droplet was spotted onto afleshly cleaved mica or silica surface Mica was used forbare DNA whereas polypeptide-coated DNA was alsoadsorbed on silica After 10min to allow for DNA adsorp-tion onto the surface the specimens were developed byimmersing them in ultrapure water for 30min followed bydrying in a stream of N2 gas

Monte Carlo simulation

In the Monte Carlo protocol the DNA chain is modelledas a string of ethN+1THORN beads which are connected by N in-extensible bonds of length lb (2829) The simulation modelconsists of bending energy between adjacent bonds and twointeraction terms The bending rigidity is set to reproducepersistence length of 50nm The interactions are hard-sphere repulsion between DNA beads and hard-wall repul-sion between the beads and the wall If the centre of a beadis beyond the channel wall the potential becomes infinitelylarge and the configuration is rejected The effectivechannel diameter is hence the real diameter minus thediameter of the bead The contour length of the DNAchain was fixed at 8mm The diameter of the bead was setequal to the bond length lb which is equivalent to the chainwidth w Accordingly for chains with a fixed contourlength of 8mm the number of beads is 1601 1068 or 801for w frac14 5 75 or 10nm respectively In each Monte Carlocycle we carried out one crankshaft and one reptationmove The simulation started from a random conformationand usually reached equilibrium after 107 cycles In theproduction run we generated 1010 cycles and recordedthe conformation every other 105 cycles For each conform-ation we calculated the extension as the maximum span ofthe DNA molecule along the channel axis Finally the ex-tension was averaged over the ensemble of 105 conform-ations Results for different persistence lengths and widthswere obtained from the scale invariance of the Monte Carloresults for the relative extension This implies that themaximum span divided by the contour length of a chainwith persistence length P width w contour length L andconfined inside a channel of diameter D is the same as theone of a chain with scaled parameters P w and L insidea channel of diameter D We have verified that effects offinite contour length on the relative stretch are unimportantfor the relevant range of channel diameters

RESULTS AND DISCUSSION

Amplification of stretch

Our nanofluidic devices are made of PDMS casted on ahigh-quality master stamp which was obtained by protonbeam writing and ultraviolet optical lithography (52324)

PAGE 3 OF 8 Nucleic Acids Research 2013 Vol 41 No 20 e189

T4-DNA (166 kb) was stained with a ratio of one YOYO-1 molecule per 100 bp With this intercalation ratio theYOYO-1 corrected contour length of T4-DNA amounts57 mm Furthermore the concomitant minimal reductionin DNA charge has no effect on the electrostatic bindingof the cationic block of the polypeptide on DNA Thestained molecules were subsequently incubated in solu-tions of various concentrations of polypeptide for atleast 24 h The molecules are accordingly coated withvarious amounts of C4K12 as indicated by the lysine toDNA NP ratio (the DNA concentration is 3mgl)In the nanofluidic experiment the coated and stained

DNA molecules were electrophoresed into an array of60 mm long and rectangular nanochannels and imagedwith fluorescence microscopy A montage of images ofsingle T4-DNA molecules in 10mM TrisHCl (29mMTrisCl 71mM Tris pH 85) and confined in channelswith a cross-section of 250 250 nm2 is shown inFigure 3A The images refer to well-equilibrated conform-ations After the electric field has been switched off themolecules relax to their equilibrium state within 60 sVideo imaging was started 5ndash10min after the moleculeswere brought into the channels and lasted for another10min With increasing C4K12 to DNA ratio theequilibrated molecules stretch in the longitudinal directionof the channel Notice that the stretch of the bottlebrushwith a 2-fold excess of polypeptide to DNA charge(NP=20) reaches 85 of the contour length Anexcess of polypeptide is needed to achieve maximumstretch owing to the dynamic equilibrium of the bindingprocess (law of mass action) A relative stretch of 85 hasalso been reported for DNA in narrow 45-nm channels aswell as for DNA at low ionic strength of 012mM in250-nm channels (617) An advantage of the present tech-nology is the less stringent conditions in terms of channeldiameter and ionic strength of the supporting mediumWe have measured the extension of the T4-DNA mol-

ecules confined in the nanochannels of various cross-sectional diameters With increasing NP ratio itbecomes more difficult to electrophorese DNA moleculesinto channels of smaller cross-sectional diameter For thehighest used NP ratio (20) we could not insert a largeenough number of molecules into channels with adiameter lt200 nm For each experimental condition wehave used a fresh PDMS replica and measured 50 mol-ecules The distributions in extension are close toGaussian Examples pertaining to different NP ratiosare shown in Figure 3B DNA fragments can easily bediscerned because their extensions fall below the valuespertaining to the intact molecules For the cut-off wehave used the mean value minus two times the standarddeviation Resolution broadening can be neglectedbecause the optical resolution of 200 nm is one order ofmagnitude smaller than the variance The relative exten-sions Lk=L that is the mean stretch divided by thecontour length of the DNA molecule are set out inFigure 4 With decreasing channel diameter andorincreasing C4K12 to DNA ratio the relative extension in-creases For NP=0 (polypeptide-free) and 01 the ex-tensions are in the range 02ndash05 times the contour lengthIn the case of NP ratios of 10 and 20 the relative

extensions are in the range 05ndash085 Furthermore the ex-tensions level off at a plateau value for smaller channeldiameters and higher NP ratios The increase in relativeextension is owing to an increase in bending rigidity andcross-sectional diameter of the bottlebrush-coated DNAThese effects will be gauged below with atomic force mi-croscopy We will however first demonstrate that large-scale genomic organization can be mapped with ourchannel platform in tandem with polypeptide coating

Large-scale genome mapping

The aim of large-scale genome mapping is not to deter-mine the genetic code at the single base level but rather toprovide a map for assembling the genome at the largerscale and to sort out large-scale variations such as inser-tions inversions and translocations For the demonstra-tion of genome mapping we have site-specifically labellednicks of -DNA (485 kb) with Alexa Fluor 546 (1517)

100D (nm)

Rela

tive

exte

nsio

n

10

02

04

06

08

200 40050

P = 240 nm w = 40 nmP = 190 nm w = 30 nmP = 95 nm w = 15 nm

P = 50 nm w = 10 nm

Figure 4 Relative extension Lk=L of T4-DNA (open symbols) and-DNA (closed symbols) in 10mM TrisHCl versus channel diameterD The NP ratios are 0 (4) 01 (5) 10 (laquo) and 20 () The solidcurves represent Monte Carlo results with noted values of persistencelength P and width w The dashed curve represents deflection theory fornarrow channels

0 10 20 30 40 500

0 1

0 2

0 3

0 4

0 5

Extension (μm)

Num

ber F

ract

ion

6 μm

NP=0 01 10 20

B

6 μm

A

Figure 3 (A) Montage of fluorescence images of T4-DNA in250 250 nm2 channels and in 10mM TrisHCl (pH 85) From leftto right NP=0 (polypeptide-free) 01 10 and 20 The scale bardenotes 6 mm and the YOYO-1 intercalation ratio is 100 bp per dyemolecule (B) Distribution in extension of a population of 50 T4-DNA molecules with indicated NP ratio Gaussian fits give mean ex-tensions of Lk=16plusmn2 25plusmn3 41plusmn3 and 48plusmn3mm for NP=001 10 and 20 respectively

e189 Nucleic Acids Research 2013 Vol 41 No 20 PAGE 4 OF 8

The YOYO-1 and Alexa labelled DNA was coated withC4K12 and electrophoresed into an array of nanochannelswith cross-sectional diameters of 150 250 nm2 From ananalysis of the YOYO-1 fluorescence we have verified thatthe relative extensions of -DNA are the same as for T4-DNA (see Figure 4) There are four resolvable Alexalabelling sites at 8 183 (average of two sites at 181 and185 kb) 313 (average of 309 312 and 318 kb) and358 kb Furthermore the sticky ends of the -DNAmolecule can also be labelled As can be seen in panel Aof Figure 5 individual Alexa labels are discernible with anNP ratio of 20 Individual labels can also be discernedfor a stoichiometric ratio of NP=10 albeit withdecreased separation In the case of polypeptide-freeDNA individual Alexa labelled sites cannot be discernedThis confirms that an excess of polypeptide to DNAcharge is needed to achieve the highest possible stretchFor comparison in panel B of Figure 5 we have displayedfluorescence images of combed bare DNA molecules Inthe combing experiment the molecules are stretched toalmost their full contour length In both the nanofluidicand combing experiments occasionally some Alexa labelsare missing (in particular the ones at the end) Out of apool of 100 molecules 2 10 and 40 of the moleculesshow 6 5 and 4 discernible labels respectively Because wedid not observe a difference in labelling efficiency forcombed (bare) and channel-confined (coated) moleculesmissing labels are due to incomplete labelling rather thancaused by the coating procedure

In the case of maximum stretch for NP=20 in150 250 nm2 channels we have analysed the Alexa fluor-escence profiles of a pool of 50 DNA molecules Anexample of the time-dependence of the Alexa fluorescenceis shown in panel C of Figure 5 On the time scale of theexperiment (90 s) the molecules are trapped inside thechannel and there is no net displacement Singlemolecule Alexa profiles were aligned and corrected for aminor variation in overall stretch of lt5 The locationsof corresponding peaks for different molecules are inregister As shown by the fluorescence images and theaverage profile in panels A and D of Figure 5 respectivelythe locations of the labels along the molecule are in goodagreement with the locations of the nicking sitesAccordingly the molecules are uniformly stretched alongthe direction of the channel without ever folding back Theaveraged Alexa fluorescence intensity profile confirms arelative stretch of 85 for NP=20 as determinedby the YOYO-1 fluorescence The resolution in peakposition is 2 kb which is similar to the one reportedfor bare DNA in 45-nm channels (17)

Persistence length and width

The persistence lengths were obtained from analysis ofatomic force microscopy images of bottlebrush-coatedDNA on silica For this purpose linearized pUC18plasmid (2686 bp) was used at a concentration of 02mgl The C4 blocks of the diblock polypeptide weakly adsorbto silica such that no additions to the buffer are necessaryto promote adhesion The C4-block mediated adsorptionof bottlebrush-coated DNA is relatively weak because it is

found that the molecules are more easily removed byflushing with water as compared with DNA moleculesthat have been adsorbed in the presence of either MgCl2or high molecular weight polylysine Accordingly thebottlebrush is not kinetically trapped in a 3D conform-ation (30) For the reference case of bare DNA moleculeswere bound to a mica surface by the use of a buffercomprising 10mM MgCl2 Excess polypeptide andMgCl2 were removed from the interface by immersion ofthe specimens in ultrapure water for 30min During thisdevelopment time the molecules equilibrate on the surfacein a 2D conformation Subsequently all specimens wereN2-dried A series of typical images for increasing C4K12

to DNA ratios are shown in Figure 6AndashD The relativelysmall plasmids are imaged in their entirety with a field ofview of 3 3 mm2 We have also imaged polypeptide-coated DNA with NP=20 However as shown inpanels E and F of Figure 6 only segments of these rela-tively long molecules can be visualized Nevertheless theirappearance in terms of thickness and rigidity is similar tothe one of the shorter plasmid moleculesTo obtain the persistence length from the atomic force

microscopy images we traced the centreline of a popula-tion of 10ndash30 individual pUC18 molecules with thesmaller number pertaining to the highest polypeptide con-centrations These were used to obtain the tangent vectorcorrelation function cos ss+L

where y is the angle

between tangent vectors at points s and s+L by averagings along the contour For weakly adsorbed DNA moleculesequilibrated in 2D conformation this correlation follows

cos ss+L

frac14 exp L=eth2PTHORNeth THORN eth1THORN

0 8 183 313 358 485 kbp

0 5 10 15Extension (μm)

NP = 20

minus5 0 5 10 15 20

40

80

Extension (μm)

C

NP = 20

0 8 183 313 358 485 kbp

D

0

8

183

313358

485kbp

0

8

183

313358

485kbp

0

B Combing

Tim

e (s

)

A NP = 20

Figure 5 (A) Montage of fluorescence images of -DNA The imagesobtained with YOYO-1 (green) and Alexa Fluor 541 (red) staining aresuperposed The molecules are confined in 150 250 nm2 channels withNP=20 Labelling sites are noted The scale bar denotes 5 mm (B) Asin panel (A) but for combed -DNA (C) Time dependence of theAlexa Fluor intensity profile pertaining to NP=20 and in a150 250 nm2 channel The exposure time is 500ms (D) AverageAlexa profile over a pool of 50 molecules with NP=20 and in150 250 nm2 channels

PAGE 5 OF 8 Nucleic Acids Research 2013 Vol 41 No 20 e189

Experimental and fitted tangent correlation functionsare shown in Figure 7 With increasing C4K12 to DNAratio fitted values of the persistence length P increasefrom 60 (bare DNA) to 240 nm at the highest polypeptideconcentration The value obtained for bare DNA agreeswith the commonly accepted value of 50 nm (31) Wehave measured the length of the molecules by tracing thecontour No change in contour length induced by thebinding of the polypeptide on DNA was observedThe atomic force microscopy images show an increase

in thickness of the bottlebrushes with increased polypep-tide concentration We analysed the cross-sectionalprofiles taken at 10 different and randomly chosen pos-itions of the molecules Gaussian fits to these profiles givesvariances of 10 15 20 and 30 nm for NP=0 01 10 and20 respectively These values should however be takenas indicative of the thickness of the brushes because theprofiles are broadened by the width of the tip and thebrushes are dried and spread out on the silica surfaceThe experimental values of the stretch can be compared

with Monte Carlo results that we have obtained for thestretching of nanoconfined semiflexible polymers varyingin persistence length P and effective thickness w The com-putational results are also displayed in Figure 4 Incomparing the experimental values to the Monte Carlosimulations we have used the values of the persistencelength P as obtained from atomic force microscopywithout any adjustment Only for bare DNA a slightlybetter fit was obtained with the nominal persistencelength of 50nm For each NP the width w used in theMonte Carlo simulations was adjusted to obtain the bestagreement with the experimental stretching data Thevalues of w obtained in this way show an increase withincreased polypeptide coating and are in reasonable agree-ment with the variances of the cross-sectional profilesobtained from atomic force microscopy

Theoretical considerations

For D lt 250 nm and NP=20 the data are close to theprediction of Odijkrsquos deflection model Here the bottle-brush is undulating along the channel by deflections fromthe wall (32) For larger values of D the data fall belowthe mark set by deflection theory owing to back-foldingand looping of the undulating wormlike bottlebrush con-formation (28)Existing theories support our findings of increasing

effective thickness and persistence lengths for bottle-brush-coated DNA (21) Assuming that saturatedbinding corresponds to full charge neutralization of theDNA phosphates by the lysines on the K12 bindingblock a single C4K12 molecule has a binding site size of6 bp and saturated binding corresponds to a distance hbetween bound polypeptides along the DNA contour of2 nm Recall that this requires an excess of polypeptide toDNA charge (NPgt 1) owing to the dynamic equilibriumof the binding process Side chain stretching inbottlebrushes is controlled by the dimensionless parameter

frac14 Nl=h eth2THORN

where N is the number of segments in a side chain l is thelength of a side chain segment and h is the distancebetween grafted side chains along the main chain If Rs

is the mean-square end-to-end distance of the stretchedside chain (in the bottlebrush) and R is the mean-squareend-to-end distance of an unstretched side chain (in freesolution) the scaling prediction for the degree of sidechain stretching takes the form (33)

Rs=R 1=4 eth3THORN

Assuming l frac14 04 nm and N=400 we find 1=4 30for the charge neutral complex with h frac14 2 nm Taking thehydrodynamic radius RH=6nm for the chain in freesolution as a measure for R (21) the scaling estimatesuggests that the side chains stretch to a value on the

20 40 60 80L (nm)

0 100

1

08

06

04

lt c

os

θ

gt

S+L

0

20

01

10

Figure 7 Orientation correlation of the tangent vectors at a pair ofpoints separated by distance L along the contour The lines representexponential fits and the polypeptide to DNA NP ratios are noted Thefitted values of the persistence lengths P=60plusmn5 95plusmn5 190plusmn10and 240plusmn10nm for NP=0 01 10 and 20 respectively

A B C

D E F

Figure 6 (AndashD) Tapping mode atomic force microscopy images oflinearized pUC18 DNA complexed with increasing amounts of poly-peptide The NP ratios are 0 01 10 and 20 in panels A B C andD respectively (E F) As in panels (AndashD) but for DNA withNP=20 The scale bar denotes 1 mm Panels BndashD are a montage

e189 Nucleic Acids Research 2013 Vol 41 No 20 PAGE 6 OF 8

order of 20 nm A precise value of Rs cannot be givenbecause the proportionality factor in Equation (3) isunknown Nevertheless a maximum cross-sectionaldiameter of the bottlebrush of 40 nm at saturatedbinding is consistent with our experimental estimate forNP=20 The persistence length P of bottlebrushpolymers can be estimated based on scaling theorywhich was verified by self-consistent field calculations(2134)

P frac14 P0+N2l3=h2 eth4THORN

with a small numerical constant frac14 002 For h frac14 2 nmand the commonly accepted value of P0=50 nm for theDNA main chain intrinsic stiffness the above theorypredicts P frac14 150 nm or a 3-fold increase as comparedwith bare DNA Hence theoretical estimates confirm theexpected order of magnitude of the stiffening effect eventhough the predicted increase is lt4- to 5-fold increaseat NP=20 that we have estimated from the analysisof the atomic force microscopy images and the extensionsof the complexes confined inside the nanochannels

CONCLUSIONS

We have shown that we can amplify the stretch ofDNA confined in nanochannels through a coating of theduplex with a diblock polypeptide For channels withcross-sectional diameters of 200 nm we can achieve auniform stretch of 85 of the contour length Owing tothe less stringent cross-sectional dimension of a fewhundred nm the channels are relatively easy to makeout of elastomer with soft lithography Furthermore theDNA molecules can be loaded within the nanochannelusing electrophoresis Monte Carlo computer simulationshows that the stretch is due to an increase in bendingrigidity and thickness of the bottlebrush-coated DNAThe persistence lengths and widths deduced from thestretch agree with what has been observed with atomicforce microscopy The highest degree of stretching corres-ponds to the deflection limit in which the bottlebrushundulates within the nanochannel without ever foldingback We have demonstrated that the amplification ofthe stretch in tandem with site-specific fluorescencelabelling allows the investigation of large-scale genomicorganization of phage DNA The limiting factor ofthe nanochannel platform is the length of the channelsWith present-day technology it is possible to fabricatelinear channels with a length of around a few millimetresso that DNA molecules of a few mega base pairs can bestretched For longer DNAs one has to use fragmentsandor nanofluidic devices with a back-folded channellayout

ACKNOWLEDGEMENTS

The authors thank Sarabjit Singh and Lingjun Zhang forexperimental support

FUNDING

Singapore-MIT Alliance for Research and TechnologySingapore National Science Foundation [CBET-0852235] Singapore MOE Academic Research Fund[MOE2009-T2-2-005] Faculty Research Grant [R-144-000-312-112] Dutch Polymer Institute project [SynProt 698] Consejo Nacional de Ciencia y TecnologiaMexico (to AHG) Funding for open access chargeNational Science Foundation

Conflict of interest statement None declared

REFERENCES

1 ReisnerW MortonKJ RiehnR WangYM YuZRosenM SturmJC ChouSY FreyE and AustinRH (2005)Statics and dynamics of single DNA molecules confined innanochannels Phys Rev Lett 94 196101

2 MannionJT RecciusCH CrossJD and CraigheadHG(2006) Conformational analysis of single DNA moleculesundergoing entropically induced motion in nanochannels BiophysJ 90 4538ndash4545

3 ReisnerW BeechJP LarsenNB FlyvbjergH KristensenAand TegenfeldtJO (2007) Nanoconfinement-enhancedconformational response of single DNA molecules to changes inionic environment Phys Rev Lett 99 058302

4 JoK DhingraDM OdijkT de PabloJJ GrahamMDRunnheimR ForrestD and SchwartzDC (2007) A single-molecule barcoding system using nanoslits for DNA analysisProc Natl Acad Sci USA 104 2673ndash2678

5 ZhangC ZhangF van KanJA and van der MaarelJRC(2008) Effects of electrostatic screening on the conformation ofsingle DNA molecules confined in a nanochannel J Chem Phys128 225109

6 KimY KimKS KounovskyKL ChangR JungGYdePabloJJ JoK and SchwartzDC (2011) Nanochannelconfinement DNA stretch approaching full contour lengthLab Chip 11 1721ndash1729

7 PerssonF UtkoP ReisnerW LarsenNB and KristensenA(2009) Confinement spectroscopy probing single DNA moleculeswith tapered nanochannels Nano Lett 9 1382ndash1385

8 SuT DasSK XiaoM and PurohitPK (2011) Transitionbetween two regimes describing internal fluctuation of DNA in ananochannel PLoS One 6 e16890

9 DaiL van der MaarelJRC and DoylePS (2012) Effect ofnanoslit confinement on the knotting probability of circularDNA ACS Macro Lett 1 732ndash736

10 ZhangC ShaoPG van KanJA and van der MaarelJRC(2009) Macromolecular crowding induced elongation andcompaction of single DNA molecules confined in a nanochannelProc Natl Acad Sci USA 106 16651ndash16656

11 JonesJJ van der MaarelJRC and DoylePS (2011) Effect ofnanochannel geometry on DNA structure in the presence ofmacromolecular crowding agent Nano Lett 11 5047ndash5053

12 ZhangC GongZY GuttulaD MalarPP van KanJADoylePS and van der MaarelJRC (2012) Nanouidiccompaction of DNA by like-charged protein J Phys Chem B116 3031ndash3036

13 RiehnR LuM WangYM LimSF CoxEC andAustinRH (2005) Restriction mapping in nanofluidic devicesProc Natl Acad Sci USA 102 10012ndash10016

14 LevySL and CraigheadHG (2010) DNA manipulation sortingand mapping in nanofluidic systems Chem Soc Rev 391133ndash1152

15 DasSK AustinMD AkanaMC DeshpandeP CaoH andXiaoM (2010) Single molecule linear analysis of DNA in nano-channel labelled with sequence specific fluorescent probes NucleicAcids Res 38 e177

16 ReisnerW LarsenNB SilahtarogluA KristensenATommerupN TegenfeldtJO and FlyvbjergH (2010)

PAGE 7 OF 8 Nucleic Acids Research 2013 Vol 41 No 20 e189

Single-molecule denaturation mapping of DNA in nanofluidicchannels Proc Natl Acad Sci USA 107 13294ndash13299

17 LamET HastieA LinC EhrlichD DasSK AustinMDDeshpandeP CaoH NagarajanN XiaoM et al (2012)Genome mapping on nanochannel arrays for structural variationanalysis and sequence assembly Nat Biotechnol 30 771ndash776

18 ReisnerW PedersenJN and AustinRH (2012) DNAConfinement in nanochannels physics and biological applicationsRep Prog Phys 75 106601

19 MinSK KimWY ChoY and KimKS (2011) Fast DNAsequencing with a graphene-based nanochannel device NatureNanotech 6 162ndash165

20 ZhangC JiangK LiuF DoylePS van KanJA and van derMaarelJRC (2013) A nanofluidic device for single moleculestudies with in situ control of environmental solution conditionsLab Chip 13 2821ndash2826

21 Hernandez-GarciaA WertenMW StuartMC de WolfFAand de VriesR (2012) Coating of single DNA molecules bygenetically engineered protein diblock copolymers Small 83491ndash3501

22 WertenMW WisselinkWH Jansen-van den BoschTJ deBruinEC and de WolfFA (2001) Secreted production of acustom-designed highly hydrophilic gelatin in Pichia pastorisProtein Eng 14 447ndash454

23 van KanJA ZhangC MalarPP and van der MaarelJRC(2012) High throughput fabrication of disposable nanofluidiclab-on-chip devices for single molecule studies Biomicrofluidics 6036502

24 van KanJA BettiolAA and WattF (2006) Proton beamwriting of three-dimensional nanostructures in hydrogensilsesquioxane Nano Lett 6 579ndash582

25 van KanJA BettiolAA and WattF (2003) Three-dimensionalnanolithography using proton beam writing Appl Phys Lett 831629ndash1631

26 ShaoPG van KanJA AnsariK BettiolAA and WattF(2007) Poly (dimethyl siloxane) micronanostructure replicationusing proton beam written masters Nucl Instrum Methods PhysRes B 260 479ndash482

27 AllemandJF BensimonD JullienL BensimonA andCroquetteV (1997) pH-dependent specific binding and combingof DNA Biophys J 73 2064ndash2070

28 DaiL NgSY DoylePS and van der MaarelJRC (2012)Conformation model of back-folding and looping of a singleDNA molecule confined inside a nanochannel ACS Macro Lett1 1046ndash1050

29 DaiL JonesJJ van der MaarelJRC and DoylePS (2012)A systematic study of DNA conformation in slit-like confinementSoft Matter 8 2972ndash2982

30 RivettiC GutholdM and BustamanteC (1996) Scanning forcemicroscopy of DNA deposited onto mica equilibrium versuskinetic trapping studied by statistical polymer chain analysisJ Mol Biol 264 919ndash932

31 WigginsPA van der HeijdenT HerreroFM SpakowitzAPhillipsR WidomJ DekkerC and NelsonPC (2006) Highflexibility of DNA on short length scales probed by atomic forcemicroscopy Nat Nanotechnol 1 137ndash141

32 OdijkT (1983) On the statistics and dynamics of confined orentangled stiff polymers Macromolecules 16 1340ndash1344

33 HsuHP PaulW and BinderK (2011) Understanding themultiple length scales describing the structure of bottle brushpolymers by Monte Carlo simulation methods Macromol TheorySimul 20 510ndash525

34 FeuzL LeermakersFAM TextorM and BorisovO (2005)Bending rigidity and induced persistence length of molecularbottle brushes a self-consistent field theory Macromolecules 388891ndash8901

e189 Nucleic Acids Research 2013 Vol 41 No 20 PAGE 8 OF 8