Nanotechnology Science and Computation Part 11 doc

We collected a large pool of SFM images about 1500 of two palindromicDNA constructs containing the curved tract of Crithidia fasciculata, bridgedhead-to-head and tail-to-tail to obtain t

254 B Samor`ı, G Zuccheri, A Scipioni, P De Santis

vature plots are obtained for each palindromic structure under investigation[18, 11, 8]

4 Experimental Evidence of DNA Sequence Recognition

by Mica Surface

Two Crithidia segments, like those in Fig 3, were ligated either in the tail (Fig 4) or in the head-to-head (Fig 5) orientation and two palindromicdimers were constructed

tail-to-Fig 4 Pictorial representation of the tail-to-tail DNA palindromic construct The

DNA monomers are composed in a palindromic fashion and sketched as curvedribbons with directionality defined by the sequence The two opposite DNA faces

are indicated with different gray intensities In (b), contrary to (a), the different

extent of the local curvature due to the differential interaction of the two monomericfaces with the surface is taken into account The dyadic symmetry is thus lost andthe C-like shape is, more correctly, expected to be G-like, instead

In 3D, the dyad axis, which characterizes the averaged shape of the dromic DNA dimers, can be oriented along any direction of space with respect

palin-to the average plane of the curved tracts This statistical symmetry constraintalso persists when the molecules are flattened on a crystal surface, such as mica

in SFM images, but only two alternative directions of the dyad axis are lowed, parallel or perpendicular to the surface plane In the former case bothcurved halves of the molecule have the same sign of the curvature sign; in thelatter case the two curved halves have curvatures opposite in sign We calledthese symmetry species C-like shape and S-like shape (or S*, the asterisk in-dicating the mirror image), respectively, because the curves are isomorphouswith these letters (Figs 4a, 5a) The C-like molecules will be characterized

al-by two positive curvatures or two negative ones, depending on which end is

Hierarchy in the Construction of DNA-Based Nanostructures 255

Fig 5 Pictorial representation of the head-to-head DNA palindromic construct.

See legend of Fig 4

chosen as the starting point of the molecule (but the sequence associated will

be the same, thanks to the palindromicity) On the contrary, the two maincurvatures will be oppositely signed in an S-like or an S*-like shape; eitherwith a positive followed by a negative one (S-shape), or by a negative fol-lowed by a positive one (S* shape), independent of the direction of which end

is chosen as the starting point of the molecule These two possibilities are theresult of the adhesion of the three-dimensional dimeric molecules on either oftheir two opposite faces In the case of C-shaped molecules (Figs 4a, 5a), thetwo faces are equivalent, instead, because within either face one half exposes

a sequence complementary to that of the other

We collected a large pool of SFM images (about 1500) of two palindromicDNA constructs containing the curved tract of Crithidia fasciculata, bridgedhead-to-head and tail-to-tail to obtain two palindromic dimers The curvaturewas evaluated along all the recorded molecular profiles and averaged over all

of them according to [18, 11, 8]

The average over the whole set of profiles however, did not vanish, asshould be expected on the basis of equal populations of the four subclassesdepicted in Figs 4a or 5a The two palindromic dimers exhibited the curvatureprofiles reported in Fig 6: the sigmoidal shapes are similar, but oppositelysigned

The non-zero curvature profiles in Fig 6 thus monitor the imbalance of thesubpopulations of the different symmetry classes This result proved that thesurface preferentially binds one of the two different faces of the curved DNAtracts (see Fig 3) and in principle can differently modify their curvature This

is pictorially illustrated in Figs 4b and 5b

We have seen that the two faces of the monomeric curved tracts exposeeither A-rich or T-rich sequences (Fig 2) By analyzing the shape assumed

by two large sets of the two palindromic dimers of the Crithidia fasciculata

256 B Samor`ı, G Zuccheri, A Scipioni, P De Santis

Fig 6 SFM ensemble average curvature profiles of the head-to-head and the

tail-to-tail DNA palindromic dimers The inversion of sequence direction results in aninversion of the sign of curvature

fragment, it turned out that the face that both dimers expose preferentially

to the mica is the T-rich one [8, 17]

5 How Effective Is This Recognition Process?

In order to answer this question we must characterize the subpopulations ofthe different classes and compare in particular that of S with that of S*

A more precise analysis of the expected shapes for the different classes hasrecently been performed [10]

We have also accounted for the presence of differential interactions tween the two faces on each monomer and the surface This is expected tochange the symmetry of the models depicted in Figs 4 and 5 In fact, suchdifferential interactions can affect the curvature and flexibility, namely the in-trinsic mechanical properties of the single monomeric units within each dimer

be-In particular, the S and S* shapes should retain the dyad axis and the sic anti-symmetry of the curvature functions but their shapes and the relatedcurvatures will no longer be quantitatively equivalent because of the differ-ent interactions with the surface Thus, the two faces of G-shaped moleculesbeing physically equivalent, they must be present on the surface in the samenumber but their contribution to the curvature will be equivalent to those of

intrin-Hierarchy in the Construction of DNA-Based Nanostructures 257

Fig 7 Predicted curvature profiles of the S,S*, G and G* symmetry classes of

tail-to-tail and head-to-head palindromic dimers as sketched in Figs 4b and 5b.Note the inversion of curvature signs of the two halves of the molecule when thetail-to-tail dimer is formally transformed in the head-to-head dimer

S+S* [10] In fact, the expected curvature profiles will be as those represented

in Fig 7 for the tail-to-tail and head-to-head dimers These predicted profileswere perfectly confirmed by those obtained experimentally (see Fig 8) byclassifying all the profiles in the four subclasses, according to their shapes,and then by averaging the curvatures plots within each subclass

The classification of the different molecules in the subclasses indicates thatthe T-rich face was deposited up to 12 times more frequently than the other.Therefore, the recognition effect is strong and the differential adhesion ofDNA to mica not only privileges one face with respect to the other but alsomodifies its curvature

258 B Samor`ı, G Zuccheri, A Scipioni, P De Santis

Fig 8 SFM ensemble average curvature profiles of the tail-to-tail palindromic DNA

dimer after shape classification The curvatures of the species are coded as in theinserted legend

6 From Statistics to Determinism

One can tailor possible applications of this recognition effect to the assembled integration of inorganic material in the construction of complexDNA-based nanostructures On the other hand, this effect has been so farcharacterized on a statistical basis only With the aim of building determin-istically designed nano-objects, rigid DNA structures must be brought intoplay These should exhibit the same segregation of complementary bases thattakes place on the two faces of a curved molecule (Fig 2) One possibility

self-is offered by the structures composed of multiple blocked Holliday junctionsdeveloped by Seeman [5, 12, 13, 14] If four blocked junctions are arranged

in a DNA parallelogram of the kind sketched in Fig 9, then the shape of theresulting object is determined precisely by the size of the arms and by thenumber of parallelograms that are assembled together in the nano-object (forinstance via sticky ends)

Within each parallelogram arm, the base sequence can be made of phasedA-tracts (see Fig 8), so that all adenines will be on one side of the par-allelogram plane, while all thymines will be on the other Since the DNAparallelogram is really a 4 nm thick object (like 2 logs of wood lying at anangle on two others) then one can design the phase of the A-tracts so that oneparallelogram can lie flat on either two A-rich DNA sides, or on two T-rich

Hierarchy in the Construction of DNA-Based Nanostructures 259

Fig 9 Parallelogram with a segregation of A and T bases on its two faces.

DNA sides Studies of the interaction of these structures with flat surfaces areunder way in our laboratories

6 D Rhodes and A Klug Helical periodicity of DNA determined by enzymedigestion Nature, 286:573–578, 1980

7 B Samor`ı and G Zuccheri DNA codes for nanoscience Angewandte ChemieInt Ed., 44:1166–1181, 2005

8 B Sampaolese, A Bergia, A Scipioni, G Zuccheri, M Savino, B Samor`ı, and

P De Santis Recognition of the DNA sequence by an inorganic crystal surface.Proc Natl Acad Sci (USA), 99:13566–13570, 2002

9 M Sarikaya, C Tamerler, A.K Jen, K Schulten, and F Baneyx Molecularbiomimetics: nanotechnology through biology Nature Mater., 2:577–585, 2003

260 B Samor`ı, G Zuccheri, A Scipioni, P De Santis

10 A Scipioni, S Pisano, A Bergia, M Savino, B Samor`ı, and P De Santis.Sequence-dependent DNA recognition by inorganic crystal surfaces Submitted,2005

11 A Scipioni, G Zuccheri, C Anselmi, B Samor`ı, and P De Santis dependent DNA curvature and flexibility from scanning force microscopy im-ages Biophys J., 83:2408–2418, 2002

Sequence-12 N.C Seeman Biochemistry and structural DNA nanotechnology: an evolvingsymbiotic relationship Biochemistry, 42:7259–7269, 2003

13 R Sha, F Liu, H Iwasaki, and N.C Seeman Parallel symmetric immobileDNA junctions as substrates for E coli RuvC Holliday junction resolvase Bio-chemistry, 41:10985–10993, 2002

14 R Sha, F Liu, and N.C Seeman Atomic force microscopic measurement of theinterdomain angle in symmetric Holliday junctions Biochemistry, 41:5950–5955,2002

15 S.J Sowerby, C.A Cohn, W.M Heckl, and N.G Holm Differential adsorption

of nucleic acid bases: relevance to the origin of life Proc Natl Acad Sci USA,98:820–822, 2001

16 S.R Whaley, D.S English, E.L Hu, P.F Barbara, and A.M Belcher tion of peptides with semiconductor binding specificity for directed nanocrystalassembly Nature, 405:665–668, 2000

Selec-17 G Zuccheri, A Bergia, A Scipioni, P De Santis, and B Samor`ı DNA on faces: adsorption, equilibration and recognition processes from a microscopist’sview In AIP Conference Proceedings DNA-Based Molecular Construction, Col-lege Park (MD) USA, p 23–27, 2002

sur-18 G Zuccheri, A Scipioni, V Cavalieri, G Gargiulo, P De Santis, and B Samor`ı.Mapping the instrinsic curvature and the flexibility along the DNA chain Proc.Natl Acad Sci USA, 98:3074–3079, 2001

Adding Functionality to DNA Arrays: the

Development of Semisynthetic DNA–Protein Conjugates

com-1 Introduction

Nature has evolved incredibly functional assemblages of proteins, nucleic acidsand other (macro)molecules to perform complicated tasks that are still daunt-ing for us to try to emulate Biologically programmed molecular recognitionprovides the basis of all natural systems, and the spontaneous self-assembly

of the ribosome from its more than 50 individual building blocks is one ofthe most fascinating examples of such a process The ribosome is a cellu-lar nanomachine, capable of synthesizing polypeptide chains using an RNAmolecule as the informational template The ribosome spontaneously self-assembles from its more than 50 individual building blocks, driven by an

∗ I wish to thank Deutsche Forschungsgemeinschaft and Fonds der Chemischen

Industrie for financially supporting our work

262 C.M Niemeyer

assortment of low-specificity, noncovalent contacts between discrete aminoacids of the protein components, interacting with distinct nucleotide basesand the phosphate backbone of the ribosomal RNA The structures of theribosomal subunits have recently been resolved at atomic resolution, and theatomic structures of these subunits and their complexes with two substrateanalogs have revealed that the ribosome is in fact a ribozyme [26] Our knowl-edge of the atomic structure of this complex biological nanomachine not onlysatisfies our desire to fundamentally understand the molecular basis of life,but it also further motivates research to emulate natural systems in order toproduce artificial devices of entirely novel functionality and performance.Biological self-assembly has stimulated biomimetic “bottom-up” approa-ches to the development of artificial nanometer-scale elements, which are re-quired commercially to produce microelectronic and micromechanical devices

of increasingly small dimensions in the range of ∼ 5 to 100 nm In this gard, Nadrian Seeman suggested early that one should fabricate syntheticnanometer-sized elements from biomolecular building blocks [60], and nowa-days DNA is being extensively used as a construction material for the fabri-cation of nanoscale systems [61] The simple A–T and G–C hydrogen-bondinginteractions allow the convenient programming of DNA receptor moieties,which are highly specific for the complementary nucleic acid Another veryattractive feature of DNA is the great mechanical rigidity of short doublehelices and its comparably high physicochemical stability Moreover, Natureprovides a comprehensive toolbox of highly specific ligases, nucleases and otherDNA-modifying enzymes, which can be used for processing and manipulatingthe DNA with atomic precision, and thus for molecular construction on thenanometer length scale

re-The generation of semisynthetic DNA–protein conjugates allows one tocombine the unique properties of DNA with an almost unlimited variety ofprotein components, which have been tailored by billions of years of evo-lution to perform highly specific functions, such as catalytic turnover, en-ergy conversion, or translocation of other components In particular, semisyn-thetic proteins conjugated with single-stranded DNA (ssDNA) oligomers, of-fer the possibility to functionalize DNA arrays with a protein content, tak-ing advantage of the specific Watson–Crick base pairing [36, 38] This chap-ter summarizes the current state of the art of the synthesis of such hybridDNA–protein conjugates and their application in DNA nanotechnology Thisapproach can be used for the self-assembly of high-affinity reagents for im-munoassays, nanoscale biosensor elements and the biomimetic “bottom-up”fabrication of nanostructured array devices

Adding Functionality to DNA Arrays 263

2 Immobilization of Proteins by Means of DNA

in a homogeneous solution, instead of in a less efficient heterogeneous phase immunosorption process Subsequently, the immunocomplexes formedcan be site-specifically captured on the microarray by nucleic acid hybridiza-tion [54]

solid-The reversibility and site selectivity of DDI enables a variety of tions, including the recovery and reconfiguration of biosensor surfaces, thefabrication of mixed arrays containing both nucleic acids and proteins forgenome and proteome research, and the generation of miniaturized biochipelements [34] Recent adaptations of DDI include the use of synthetic DNAanalogs, namely pyranosyl–RNA oligomers, as recognition elements for theaddressable immobilization of antibodies and peptides [70] and the DNA-directed immobilization of hapten groups for the immunosensing of pesticides[4] Recently, the DDI method has been applied in functional genomics toidentify the members of a small-molecule split-pool library which bind to pro-tein targets [73, 72, 17, 11, 71] In this approach, libraries of peptide ligandsare encoded by peptide–nucleic acid (PNA) tags After the library has beenincubated with a mixture of potential binding proteins, the PNA tags arethen used for the deconvolution of the library using DNA microarrays Theseapproaches have recently been reviewed elsewhere [27]

applica-DDI has also been applied to inorganic gold nanoparticles thereby abling the highly sensitive detection of nucleic acids in a DNA microarray

en-264 C.M Niemeyer

format [65, 56] This approach has opened the door to a large number of mercially relevant applications in the area of bioanalytics, which have recentlybeen reviewed elsewhere [59] With respect to further miniaturization, Demers

com-et al have adapted the DDI approach to surface nanostructuring by “dip-pen”nanolithography (DPN) [10] DPN employs an SFM tip to “write” thiolatedcompounds with less than 30 nm linewidth resolution on gold substrates [15].The direct writing of thiol- and acrylamide-modified oligonucleotides [9] al-lows the production of nanostructured DNA arrays, which can be used as animmobilization matrix for use in DDI

Nanostructuring based on the self-assembly of DNA junctions and tilemotifs might offer an even more powerful way to fabricate high-densitiy pro-tein arrays (Fig 1b) Initial attempts in this direction took advantage of 2Dnanogrids, consisting of four-arm junctions, which contained regular arrange-ments of biotinyl groups Incubation with the biotin-binding protein strepta-vidin (STV) led to the formation of periodic protein arrays [74, 55] Theseexamples demonstrate that a DNA nanoarchitecture can be used as a scaffoldfor arranging proteins, and it will be straightforward to extend this approach

to DNA grids containing programmable sites for protein immobilization Inthis respect, DDI-based techniques will be an option for fabricating nanoscalefunctional devices composed of DNA grids which are decorated with severaldifferent proteins

(a)

(b)

Fig 1 (a) Schematic drawing of the DNA-directed immobilization (DDI) method.

A microrarray of captured oligonucleotides is used as an immobilization matrix forsite-specific binding of proteins tagged with complementary nucleic acid oligomers.Note that owing to the specificity of Watson–Crick base pairing, many differentcompounds can be site-specifically immobilized simultaneously in a single step Re-

produced from reference [27], with kind permission; (b) Illustration of a possible

implementation of the DDI method to functionalize DNA nanoarrays with proteins

Adding Functionality to DNA Arrays 265

3 Functional Multiprotein Assemblies

The concept of using DNA as a framework for the precise spatial ment of molecular components, initially suggested by Ned Seeman [60], wasexperimentally demonstrated by positioning several of the covalent DNA–STV

arrange-conjugates 4 along a single-stranded nucleic acid carrier molecule containing

a set of complementary sequences (Fig 2) [51] The covalent conjugates 4

can be used as versatile molecular adaptors because the covalent attachment

of an oligonucleotide moiety to the STV provides a specific recognition main for a complementary nucleic acid sequence in addition to the four nativebiotin-binding sites For instance, supramolecular DNA–protein nanostruc-

do-tures (e.g., 5 in Fig 2) have been assembled as model systems to investigate

the basic principles of the DNA-directed assembly of proteins [51, 47] Thesestudies showed that, in particular, the formation of intramolecular secondarystructures in the nucleic acid components often interferes with the effectiveintermolecular formation of the supramolecular DNA–protein assemblies [33].The DNA-directed assembly of proteins can be applied to fabricate ar-tificial, spatially well-defined multienzyme constructs, which are not accessi-ble by conventional chemical crosslinking In biological systems, multienzymecomplexes have mechanistic advantages during the multistep catalytic trans-formation of a substrate because reactions limited by the rate of diffusionaltransport are accelerated by the immediate proximity of the catalytic centers.Furthermore, the “substrate channeling” of intermediate products avoids side

reactions As an example, STV conjugates 4 were used to assemble bound bienzymic complexes, such as 8 in Fig 2, from biotinylated luciferase

surface-and oxidoreductase [49] The total enzymatic activities of the tase/luciferase bienzymic complexes, which catalyze the consecutive reactions

oxidoreduc-of flavinmononucleotide reduction and aldehyde oxidation, depended on theabsolute and relative spatial orientation of the two enzymes Not only are suchstudies useful for exploring proximity effects in biochemical pathways, but alsothe investigation of artificial multienzymes will allow the development of novelcatalysts for enzyme process technology, capable of regenerating cofactors or

of performing multistep chemical transformations of cheap precursors intodrugs and fine chemicals

With respect to synthetic nanosystems and materials science, the opments in the DNA-directed organization of semiconductor and metal nan-oclusters [59, 63, 35, 22, 7, 52] have stimulated the application of DNA–STV

devel-conjugates 4 to organizing biotinylated gold nanoclusters to generate novel biometallic nanostructures, such as 6 in Fig 2 [47] Given that the conjugates

4can be used like components of a molecular construction kit, functional teins, such as immunoglobulins, can be conveniently incorporated into thesebiometallic nanostructures A proof of feasibility was achieved by the assem-

pro-bly of the IgG-containing construct 7 (Fig 2), capable of specifically binding

to surface-immobilized complementary antigens [47]