Thomas J.
Bandy
,
Ashley
Brewer
,
Jonathan R.
Burns
,
Gabriella
Marth
,
ThaoNguyen
Nguyen
and
Eugen
Stulz
*
University of Southampton, School of Chemistry, Highfield, Southampton SO17 1BJ, UK. E-mail: est@soton.ac.uk; Fax: +44 (0)23 80 59 68 05; Tel: +44 (0)23 80 59 93 69
First published on 9th August 2010
Oligonucleotides have recently gained increased attraction as a supramolecular scaffold for the design and synthesis of functional molecules on the nanometre scale. This tutorial review focuses on the recent progress in this highly active field of research with an emphasis on covalent modifications of DNA; non-covalent interactions of DNA with molecules such as groove binders or intercalators are not part of this review. Both terminal and internal modifications are covered, and the various points of attachment (nucleobase, sugar moiety or phosphodiester backbone) are compared. Using selected examples of the recent literature, the diversity of the functionalities that have been incorporated into DNA strands is discussed.
From left: Jonathan R. Burns, Ashley Brewer, ThaoNguyen Nguyen, Eugen Stulz, Gabriella Marth and Thomas J. Bandy | Thomas J. Bandy completed his undergraduate studies at the University of Southampton, submitting his Masters thesis on chiral Ruthenium and Osmium complexes under the supervision of Prof. F. Richard Keene at James Cook University, Queensland, before receiving his Masters degree in chemistry in 2007. He then joined the research group of Dr Eugen Stulz at the University of Southampton, and is currently in the final year of his PhD where his research interests include the templated assembly of supramolecular arrays of fluorophores and circular dichroism. |
Ashley Brewer studied as an undergraduate at the University of Southampton and worked on undergraduate research projects under the supervision of Prof. Fred Wudl at UCLA and UCSB; he obtained his Masters in Chemistry in 2007. He joined Dr Eugen Stulz's research group at the University of Southampton in 2007 and is currently a final year PhD student working on porphyrin substituted DNA arrays for use as supramolecular wires. |
Jonathan R. Burns in 2007 received a degree in chemistry at the University of Southampton which involved undergraduate research with Prof. Tom Brown. Currently a final year PhD student at Southampton University working with Dr Eugen Stulz on energy transfer between porphyrins. |
Gabriella Marth attended the University of Technology and Economy in Budapest, Hungary, where she graduated in chemical engineering in 2005. She received her PhD degree in 2008 at Sunderland University for her work on the synthesis of polyfunctional pyrroles and investigation of the chemoselectivity of their reactions. After completing her doctorate she stayed at Sunderland University for a postdoctoral fellowship to develop a new synthetic route of natural product analogues under the supervision of Prof. Paul W. Groundwater and Prof. Rosaleen Anderson. In 2009, she joined Dr Eugen Stulz's group as a postdoctoral researcher working on a software-controlled assembly of oligomers. |
ThaoNguyen Nguyen studied Chemistry at University College, Oxford University, and completed her M. Chem degree under Dr Josephine M. Peach's supervision. ThaoNguyen is currently finishing her PhD thesis, titled “Porphyrin–DNA as Scaffold in Nanotechnology”, in the group of Dr Eugen Stulz. |
Eugen Stulz received his PhD degree from the University of Bern, Switzerland, for studies in the field of artificial nucleases (Prof. Christian Leumann), and moved to Cambridge, UK, as postdoctoral fellow in 1999 to work in supramolecular porphyrin chemistry (Prof. Jeremy K. M. Sanders). In 2003 he moved to Basel where he held an independent position as Fellow of the Treubel Fonds (Habilitation). In 2006 he was appointed lecturer at the University of Southampton and was promoted senior lecturer in 2010. His research interests are in self-assembly of molecular systems based on (bio)molecules, synthesis of nano-materials, and their applications in electronics and medicine. |
Fig. 1 Left: structure of B-type DNA duplex; right: A–T and G–C base-pairs and numbered structures of the four nucleosides, deoxyadenosine, thymidine, deoxycytidine and deoxyguanosine. |
The reliability with which a specific DNA sequence recognises its complementary sequence has been used in the past ten years or so to create new nanometre-scale two- and three-dimensional objects such as grids and lattices on surfaces,4,5 nanoscale patterns through folded DNA,6 bipyramids,7 cubes and cages,8 all based on native DNA strands.9 Also RNA has been used as a building block in nanotechnology.10 Nano-structures using chemical modifications to introduce additional points of connectivity adds further to the repertoire of available geometries.11,12 Two-dimensional DNA structures on surfaces can be used to further position proteins or nanoparticles in order to form grids that may be of use in diagnostics.4,13 The commercial availability of strand modifiers, in particular thiol end-modifiers, allows for easy attachment of DNA strands onto gold nanoparticles (AuNPs). With these techniques, different AuNPs have been connected site-specifically using complementary DNA strands attached to the particles.14 The utility of novel DNA architectures in supramolecular chemistry, biology and nano-technology clearly demonstrates the versatility of this approach.15,16 The combination of DNA modification with a programmed architecture will certainly be one of the future strengths of this approach, and complementing similar efforts undertaken using peptide- or carbohydrate-based scaffolds. Applications of functionalised DNA have recently been reviewed separately, in particular, novel strategies for site-specific DNA labelling,17chromophore labelling for photoactive DNA-based nanomaterials,18assembly of chromophores guided by nucleic acids19 and biological applications of conformationally restricted DNA analogues.20
The introduction of modifications onto DNA can be achieved via several different methods at various positions. Sites available for modification include: 3′- and 5′-terminal positions; 2′- and 4′-positions to the ribose ring; and finally, modifications to one of the four natural bases, A, C, G, and T. In addition, the nucleobase itself can be substituted with designer molecules (artificial nucleobases), or the entire nucleotide can be replaced with moieties that mimic the function and structure of the DNA (base surrogates). These modifications are discussed in the following sections in more detail. Since a complete coverage of the literature in this rapidly growing field of research is far beyond the scope of this tutorial review, illustrative examples are selected to demonstrate the versatility of the individual approaches.
In general, design of the modification and choosing an appropriate methodology highly depends on strategic aims. If the modification is to be attached via automated DNA synthesis, then the building block has to be compatible with standard DNA synthesis chemistry. In particular, compatibility with the deprotection of the 4,4′-dimethoxy trityl protection group (3% TCA in DCM), oxidation (iodine solution), and cleavage from the resin (conc. aq. NH4OH) may be limiting factors. For 5′-end modification, the protecting group can be changed to avoid need for acid treatment. Post-synthetic modification is an alternative strategy, which avoids these issues, but may be restricted to a limited solvent range, such as aqueous buffers, methanol, DMSO or DMF. Another efficient method is use of PCR for the construction of longer DNA duplexes than would be accessible by standard solid-supported synthesis if the polymerase accepts the modification, although producing low quantities of DNA. This method requires the synthesis of 5′-triphosphate nucleotides but does avoid the use of reactive phosphoramidite monomers. The major drawback of this methodology is that there is no control over the site of incorporation of the modification in longer and more complex DNA sequences.
Fig. 2 The top part shows a selection of commercially available end-modifiers. CPG = controlled pore glass, DMT = 4,4′-dimethoxy trityl, PG = protecting group. Bottom: 5′-modification through tailor-made phosphoramidites (left) or substituted nucleotides (right). |
The second method links the modification via direct functionalisation of the 5′-position of the ribose moiety prior to coupling of the modified monomer. The chemistry may be more flexible compared to the phosphitylation of modifiers. The 5′-position of thymidine, for example, is particularly easily functionalised as protection of the nucleobase is not normally required (as opposed to A, C and G). Examples include ester formation, oxidation of the 5′-OH to the aldehyde for use in reductive amination, and transformation of the 5′-OH to an amine for amide formation. End-modification also has the advantage that the DNA is much more tolerant, in terms of structure and properties, of these substituents as there are no steric constraints compared to internal modifications (see below).
5′-Terminal modifications have been studied for a variety of reasons, but primarily to increase stability towards enzymatic degradation, to increase thermal stability of the duplex through capping, to enhance target affinity facilitating detection (labelling) or for monitoring structural changes (Fig. 3). One of the most comprehensive studies in terms of end-capping substituents was reported by Richert et al., who screened a total of 52 modifications, ranging from glycine to vancomycin.21 Berova et al. have used porphyrins as circular dichroism (CD) markers to monitor structural changes in DNA (Fig. 3).22 The attachment of this achiral chromophore to DNA places it into a chiral environment, and allows excitonic coupling between porphyrin moieties giving rise to characteristic CD spectra. These are very sensitive to the porphyrin environment,23 thus allowing, for example, detection of the change from B to Z form of DNA.22 It should be noted that in addition to UV-Vis spectroscopy, CD spectroscopy is a major tool for analysis as it provides very useful information about the DNA's global structure.
Fig. 3 Principle of end-capping for increased DNA stability (left),21 and attachment of a porphyrin marker for induced CD spectroscopy (right).22 |
Selective attachment of metals to the end of the DNA has been achieved through ligand binding, allowing the construction of metal templated frameworks and supramolecular assemblies.24,25 By careful control of ligand design, different metals can be selectively bound to the DNA duplex and used, for example, to study charge transport phenomena.26 Using this approach, Sleiman et al.27,28 were able to bind a range of 3d transition metals selectively between two duplexes containing either terpyridine or phenanthroline ligands (Fig. 4). For example, spontaneous oxidation of Cu(I) to Cu(II) was observed when the metal ion was placed in the incorrect environment for its optimum geometry, which was reversed when the coordination environment was altered back to the matching geometry. The site-specific incorporation of metal centres into the DNA duplex also dramatically raises the melting temperature with ΔTM = 43 °C.
Fig. 4 Terpyridine and phenanthroline ligands used to alter the coordination environment on the 5′-terminus, and to selectively and controllably bind specific metals to the duplex.28 Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission. |
Fig. 5 3′-End monothio phosphate, which can act as a nucleophile (top left);302′-amino substitution for post-synthetic functionalisation through imine-formation (top right),313′-amino modified DNA for introduction of carboxylic acid substituted functional groups (bottom).32 |
Meade et al.31 have reported a method for attaching a modification to the 3′-end of the DNA strand by attachment to the 2′-position of the ribose moiety. The 2′-position of a ribose or deoxyribose ring is available for modification which, if the modified monomer is used as the first base in a 3′- to 5′-sequence on a universal support, will place the modification at the 3′-end of the strand (Fig. 5). This approach allows the 3′-end to carry out its role in the 3′- to 5′-DNA synthesis, thus avoiding the limitations of post DNA synthesis modification. Meade's approach was to condense a 2′-amino nucleoside and an aldehyde to yield a 2′-imino modified nucleoside. The resulting modification was a bidentate ligand (through nitrogen lone pairs) which allowed for metallation by a ruthenium bipyridine complex.
An alternative synthetic pathway to 3′-modification was demonstrated by Ihara et al.32 In this case, 3′-amino modified DNA was synthesised by standard methods, and the strands were deprotected and cleaved from the solid support before coupling to an anthracene carboxylic acid for photo-induced chemical ligation of DNA strands (Fig. 5). This post-DNA synthesis coupling could be applied to any carboxylic acid containing moiety, thus avoiding occasional problems during phosphitylation and coupling of modified nucleobases. In fact, direct addition of commercially available modifiers such as amino, thiol and carboxy modifiers to a universal support leads to the same methodology as in post-synthetic 5′-end modification and is usually the method of choice if modifications are to be incorporated here.
Fig. 6 Examples of artificial nucleosides33 and schematic representation of zipper-like inter-strand stacking within the DNA duplex. |
A large variety of nucleosides with artificial nucleobases have been synthesised, including but not limited to phenanthroline, naphthalene, stilbene, pyrene, coumarin, terphenyl, biphenyl, bipyridine and porphyrin, most of which have been covered in a review by Kool.33 Examples of conventional base analogues are also included in this review. The aromatic nature of the artificial nucleobases is crucial for obtaining a stable duplex through π-stacking, although the dsDNA is remarkably flexible, tolerating and accommodating artificial bases that are not perfect mimics of the natural bases. The thermal analysis and structural solution by NMR spectroscopy of multiple bi-phenyl modified dsDNA by Leumann demonstrates for the first time that interstrand stacking can increase duplex stability,34 and hydrogen bonding between the natural bases is not crucial in these systems.35 Multiple insertions of these base-surrogates lead to a zipper-like arrangement (Fig. 6).36 Such modifications are very attractive as fluorescence markers; the fluorescence is normally quenched to a great extent when encapsuled in a perfect DNA duplex, but is retained to a large portion in a non-ideal duplex environment, e.g. when bulges or mismatches are present in the flanking sequences. A remarkable exception is a binaphthyl nucleoside recently reported by Seitz,37 which shows an increase in fluorescence upon multiple incorporation into DNA.
Other more recent advances by Kool et al.38 describe mimicking the conventional base-pairs using expanded ring structures, denoted yDNA bases (Fig. 7). Despite their increased size, the novel nucleobases can be recognised and replicated in vivo by natural polymerases, and have also been successfully replicated within bacterial cells. This example also very nicely demonstrates that the use of artificial nucleobases is not only restricted to the synthetic laboratory for use in supramolecular chemistry, but may well have applications in synthetic biology for expanding the genetic alphabet.
Fig. 7 Comparison between natural Watson–Crick base-pairs A·T and G·C (top), and the analogous ‘yDNA’ base pairs A·yT and G·yC (bottom).38 |
Examples of metal-controlled “base-pairs” include the pyridine-2,6-dicarboxylate and pyridine-2,6-dicarboxamide nucleobases, each of which form a Cu(II)-mediated base pair with a pyridine nucleobase on different strands (Fig. 8a). By replacing a single base pair in the middle of a 15-mer sequence, Schultz et al.39 have demonstrated an increased stability of the duplex which is manifested by an increase in the melting temperature Tm compared to natural DNA.
Fig. 8 (a) A Cu(II) mediated metallo-base-pair.39 (b) Schematic representations of Cu(II) and Hg(II) mediated metallo-base-pairs.41 The metal ions reside within the double helix itself, and the sequence can be precisely controlled during the DNA synthesis.‡ Reprinted by permission from Macmillan Publishers Ltd:41 copyright 2006. |
An extension of the concept of using artificial metal-binding nucleobase is hydroxypyridone as ligand.40 This initial system was expanded by Shionoya and Carell41 to engineer a system including the hydroxypyridone nucleobase as one of two different metal-complexing units (alongside a pyridyl nucleobase), to selectively produce an oligonucleotide double helix with a string of complexed metal ions in replacement of the standard base pairs (Fig. 8b). Two hydroxypyridone units on adjacent strands form a planar “base pair” on complexation with Cu(II) ions, whilst two pyridyl units complex an Hg2+ ion. It was also demonstrated that thymine–thymine mismatches can selectively bind Hg(II) ions in the system, with standard DNA as flanking sequences to aid preorganisation. Cu(II) and Hg(II) ions can then be selectively bound in the desired sequence dictated by the sequence of cation binders, leading to a programmed assembly of a hetero-metallic nanowire. The system was recently expanded to self-assemble a triplex around octahedrally coordinated Fe(III) ions.42 A triplex containing four Fe(III) base triplets was synthesised, and so demonstrates the extension of this technique to allow octahedrally-coordinated transition metals to be incorporated into metal–oligonucleotide complexes and nanowires.
Häner et al.43 have synthesised an oligomeric DNA strand with a central section consisting of up to seven pyrene modified subunits (Fig. 9). The system is self-organising and forms a duplex with the formation of an interstrand helical stack of pyrene subunits. Despite the fact that pyrene subunits are planar and achiral themselves, they adopt a right-handed helix which is imposed on the system by flanking natural DNA sequences. Other mixed DNA hybrids have been synthesised using tetrathiafulvalene, perylene diimide and phenanthrene modifications.44
Fig. 9 Schematic representation of oligopyrene repeat units flanked by DNA double-helix sequences.43 Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission. |
Analogously, the ribose can be replaced with a short alkyl chain having a perylene unit on a side chain as described by Wagenknecht et al.45 The short alkyl chain mimics the distance between adjacent phosphates in natural oligonucleotides (Fig. 10). Each modified perylene unit is placed opposite an abasic site on the adjacent strand, and the perylenes are positioned alternately to form a zipper-like interstrand sequence. Again, this adopts a right-handed helix with the presence of natural Watson–Crick base pairs at both termini. By this method, organic aromatic molecules can be arranged into a helical array where the supramolecular structure is predetermined by the DNA.
Fig. 10 Structure of perylene subunit (left) as base surrogate; schematic view of DNA double helix (middle) with central nucleobases replaced with alternating perylene moieties;45 ligated DNA strand obtained by DNA templated metal compex formation (right).29 Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission. |
The backbone modification described by Sheppard29 includes the synthesis of two natural oligonucleotide strands, one with salicylaldehyde units at both the 3′- and the 5′-terminal position. Addition of the complementary strand aligned the two salicylaldehyde units in adjacent positions, and addition of a square planar metal ion, such as Mn(II) or Ni(II), was used to ligate the two sequences (Fig. 10). The metal ion is removed and the result is two strands ligated about a central spacer. Such a system could be expanded to allow facile connection of multiple oligonucleotide strands; however, it is metal-templated DNA synthesis that is very promising for the synthesis of novel biomolecules. This system has recently been extended by Brown46 to the use of copper-catalysed azide–alkyne cycloaddition (CuAAC) reaction which is the best example of click-chemistry. The tolerance of the triazole linkage by polymerases suggests great potential of this system in bio-organic chemistry.
All nucleobases are commercially available as their iodinated analogue. However, by far the most commonly used starting nucleobase is 5-iodo-2′-deoxy uridine (5-I-dU), primarily because thymidine (and dU) do not normally require protection of the nucleobase, and the 5-I-dU is the most cost-effective starting material. Modification of iodinated bases by Sonogashira cross-coupling53 or by Stille cross-coupling54 is facile and allows for various different substituents to be attached to the DNA. The majority of modifications are attached to monomers prior to DNA synthesis. Alternatively, Richert has demonstrated that Sonogashira coupling to an iodinated single-strand whilst on the solid support is possible for a wide variety of substituents.52
If the site of functionalisation on the nucleobase is chosen appropriately, the modification will protrude from the double helix into the major groove, whilst the base pairs still interact via Watson–Crick hydrogen bonding. This geometry does not seem to distort the helix to a large extent. To maintain the helical structure and the base pairing, pyrimidines should be modified at the 5-position, and purines should be replaced with 7-deazapurines with modifications made at the 7-position. An example by Famulok et al.55 shows that all nucleotides in a DNA strand can be modified with functional groups which are derived from amino-acid side chains (Fig. 11). These are suitable substrates for polymerases, thus a fully functionalised DNA can be amplified via PCR. That even larger substituents are tolerated by polymerases was shown early on by Seela.56
Fig. 11 Modified nucleoside-5′-O-triphosphates as substrates for PCR mediated DNA synthesis of fully modified functional DNA (top)55 and of metal complex functionalised DNA (bottom).50 |
The use of cross-coupling reactions of nucleoside triphosphates and consequent use in PCR has been reviewed recently by Fojta and Hocek.57 Their work also confirmed that modification at the 7-position of a 7-deaza purine-5′-O-triphosphate is preferred over the 8-position. This is because the latter provide poor substrates for DNA polymerases due to enhanced steric hindrance in the DNA backbone.50 The cross-coupling approach involves direct attachment of metallated [Ru(bpy)3]2+ and [Os(bpy)3]2+ acetylenes onto the nucleobase-triphosphates (Fig. 11). The Ru(II) and Os(II) bipyridine metal complexes were attached to all four bases, and the building blocks were incorporated into DNA using PCR to enable sequence-specific incorporation of the metal complexes.56
Ru(II) and Os(II) based chromophores have also been used for photophysical studies whereby the chromophores exhibit donor- and acceptor-type interactions, respectively. In particular, Tor et al.58 have attached [(bpy)2Ru-3-ethynyl-1,10-phen]2+ and [(bpy)2Os-3-ethynyl-1,10-phen]2+ chromophores rigidly via an acetylene bond to dU. Systematic variation of the distance-separation between chromophores revealed quenching of the Ru-complex's fluorescence which was approximately proportional to the Förster dipole–dipole mechanism (Fig. 12). Notably, they also discovered that changing the rigid acetylene linker to a flexible dimethylene linker, the donor–acceptor pairs revealed a greater correlation with Förster type behaviour. Synthesis of these strands was achieved using standard solid-phase DNA techniques, analogously to what was reported by Grinstaff51 for bipy-complexes attached via a propargyl-amide linker. This also demonstrates that the use of solid-phase DNA synthesis is a better method to control precise positioning of functionalities than PCR.
Fig. 12 Base-pair separation of chromopores vs. FRET efficiency based on emission quenching of a ruthenium complex. Reprinted with permission from ref. 58. Copyright 2002 American Chemical Society. |
The examples which have been described above demonstrate how DNA can be modified externally by different chemical functionalities, upon which a variety of chromophores can be attached which may potentially have applications in redox and fluorescence labeling, drug delivery and nanotechnology; the latter potential was laid out in a feature article by Wengel.59 Indeed it is this application that has attracted major interest, and DNA is being used as a scaffold to create photonic and electronic wires. Recent work by Wagenknecht,48 Stulz,60 Wengel,61 Schuster62 and Nakamura and Yamana63 showed that fluorophores can be covalently connected to DNA and RNA and aligned within the grooves to form a helical array of stacked chromophores (Fig. 13). Again, most commonly used sites for modification are the 5-position of dU48,60 or C,62 the 2′-hydroxy group of the ribose,64 or 2′-amino modified LNA building blocks.61 It should be noted that attachment of the modifications to the 2′-position of the ribose will direct the modifications into the minor groove of the DNA, as compared to attachment onto the nucleobase.
Fig. 13 Comparison of the calculated structures of pyrene–DNA,49porphyrin–DNA,65pyrene–RNA,63benzoyl-LNA61 and aniline–DNA62 (from left to right).§ |
The stability of the resulting array depends on the design of the system, i.e. if the chromophores are attached to one strand only, or if they are attached to both complementary strands in an alternating manner. The latter leads to external zipper-arrays (see also 3.1 for the discussion on internal zipper-arrays). For example, zipper-like stacking of pyrenes does not greatly reduce the stability of the duplex, and it is certainly an advantage if the integrity of the system is maintained.48 Hydrophobic substituents such as porphyrins tend to destabilise the DNA duplex significantly if attached to one strand only, despite the indication that the structure is not greatly altered as shown by CD spectroscopy and molecular modelling.47 However, attachment of porphyrins onto complementary strands forms a zipper-like arrangement with enhanced duplex stability, most likely due to π–π-stacking and hydrophobic interactions. The porphyrins have the advantage that they can be metallated without disturbing the dsDNA, thus the reversible formation of potential photonic wires based on metal complexes becomes possible.65
Another strategy to efficiently assemble porphyrin–DNA structures was followed by Seeman and Majima. Herein a maleimide substituted tetraphenyl porphyrin was conjugated with four thiol functionalised DNA strands. Upon duplex formation with complementary strands, four double helices were assembled and used to create porphyrin diads, showing efficient energy transfer from a Zn porphyrin to a free base porphyrin (Fig. 14a).66 Potentially, this DNA structure could accommodate host–guest system within the cofacial porphyrin dimers. This four-way-branched DNA was also used as a connector in a DNA tile system to build DNA nano-tubes (Fig. 14b).67 These examples illustrate that DNA as a supramolecular scaffold provides an excellent backbone for external covalently linked functionalities, ranging from optoelectronically active chromophores to organic polymers such as nylon, as shown by Seeman who described a nylon/DNA ladder polymer.68
Fig. 14 (a) Four-way-branched porphyrin–DNA assembly (Reprinted with permission from ref. 66. Copyright 2005 American Chemical Society); (b) use of porphyrin–DNA in a tile system to create DNA tubes (Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission).67 |
In general, chromophores show strong electronic interactions when attached to DNA, and energy transfer can be achieved along the DNA with the appropriate design of the array, as shown by Tinnefeld.69 A recent example by Brown and Norden further demonstrates the suitability of DNA based nano-architectures for the creation of artificial photosynthetic systems.70 In this example, a DNA based assembly of a light-absorbing antenna (fluorescein) and a redox switch (porphyrin) were anchored onto a lipid membrane. Light-induced energy transfer from the fluorescein to the porphyrin triggered electron transfer from the porphyrin to a membrane based quinone derivative. The excitation energy is therefore trapped in the lipid phase of the membrane in the form of a radical anion, which might be used for further chemical reactions.
Footnotes |
† The standard controlled pore glass (CPG) solid supports have the first nucleotide already attached to the beads. A universal support comes “unloaded”, i.e. there are no nucleotides or other modifiers attached to the CPG, thus any phosphoramidite that is loaded onto the universal support will inevitably form the 3′-end of the DNA strand. |
‡ We believe that the right-hand DNA strands depict the incorrect directionality of the DNA. The figures were redrawn or reprinted from the corresponding publications. |
§ (a), (b) Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission. (c), (d) Reproduced by permission of the Royal Society of Chemistry. (e) Reprinted with permission from ref. 62. Copyright 2008 American Chemical Society. |
This journal is © The Royal Society of Chemistry 2011 |