Nanoscale Nucleic Acid Recognition at the SolidâLiquid Interface

Subscriber access provided by University of Sunderland

Invited Feature Article

Nanoscale Nucleic Acid Recognition at Solidliquid Interface using Xeno Nucleic Acid Probes HIYA LAHIRI, Sourav Mishra, and Rupa Mukhopadhyay Langmuir, Just Accepted Manuscript • DOI: 10.1021/acs.langmuir.8b02770 • Publication Date (Web): 06 Nov 2018 Downloaded from http://pubs.acs.org on November 7, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 45 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Nanoscale Nucleic Acid Recognition at Solid-liquid Interface using Xeno Nucleic Acid Probes Hiya Lahiri, Sourav Mishra† and Rupa Mukhopadhyay* School of Biological Sciences, Indian Association for the Cultivation of Science, Jadavpur, Kolkata 700 032, India

Abstract Challenges in reliable nucleic acid detection are manifold. The major ones are related to false positive or negative signals due to a lack of target specificity in detection, and to low sensitivity, especially when a plethora of background sequences are present that can mask the specific recognition signal. Utilizing designed synthetic nucleic acids that are commonly called xeno nucleic acids could offer potential routes to meet such challenges. In this article, we have presented the general framework of nucleic acid detection, especially keeping nanoscale applications in mind, and discussed how and why the xeno nucleic acids could be truly an alternative to the DNA probes. Two specific cases of locked nucleic acid (LNA) and peptide nucleic acid (PNA), which are nuclease-resistant and can form thermally stable duplexes with DNA, have been addressed. It is shown that the relative ease of the conformationally rigid LNA probe to be oriented upright on substrate surface, and of the non-ionic PNA probe to result into high probe density assist in their use in nanoscale nucleic acid recognition. It is anticipated that success with these probes may lead to important developments such as PCR-independent approaches where the major aim is to detect low number of target sequences in the analyte medium.

† Present address: Department of Chemistry, Pohang University of Science and Technology (POSTECH), 77 Cheongam-Ro, Pohang, Republic of Korea 790-784. * Corresponding author: Dr. Rupa Mukhopadhyay (Telephone: +91 33 2473 4971 Extn. 1506; Fax: +91 33 2473 2805; E-mail: [email protected]) 1 ACS Paragon Plus Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 45

Introduction In modern times, the human living condition has been considerably benefitted by the use of biosensors, widely encompassing the cases of clinical diagnostics, food inspection, environmental monitoring and forensic detection.1-3 Being an analytical device, a properly engineered biosensor is expected to provide signals that are target-specific and sensitive to low target concentrations, i.e., nanomolar (10-9 M) to attomolar (10-18 M), depending on the type of the biomolecular analyte. This level of sensitivity is often difficult to achieve in case of macro/micro scale biosensors, since such sensors employ significantly large surface areas and/or reaction volumes that capture multiple sensing events and require a sizeable number of analyte molecules for a reliable signal to build up. The nanoscale science and technologies, as developed over the last three decades, are now poised to address this issue of sensitivity, since considerable advancements have been made in fabrication of nanometer length scale sensing surfaces4,5 and in electronic circuitry to convert very weak signals (pA to fA order) into a measurable electronic readout.6 It has been shown recently that recognition event can be ‘molecularly resolved’ using a ‘nm’ length scale sensing surface,7 and that a sensing surface can be integrated in nanoelectronics-based circuitry that allows readout of weak signals.8 This capacity of a nanobiosensor to respond to low amount of target is a paradigm shift from the traditional approaches like gel electrophoresis, where PCR-based target pre-amplification is essential. Apart from the ability of detecting low concentrations of analyte, a precise and sensitive biosensor also allows monitoring of biological interactions, which are generally weak in nature, and involve very low forces (pN order);9 and/or angstrom to nanometer length scale changes in the molecular conformation/assembly structure.10 For continuous monitoring of fluctuations of analyte concentration in body fluid, where the sensor is attached either onto the skin surface or immediately underneath, involvement of nanoscale strategies offer encouraging possibilities.11 2 ACS Paragon Plus Environment

Page 3 of 45 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Since genes are the precursors of proteins, study of gene polymorphisms and gene mutations play an important role in early and accurate disease diagnosis and in prediction of disease susceptibility. Biosensing of disease-relevant single nucleotide polymorphisms (SNPs), which are an important class of biomarkers, has opened up a key route to person-specific therapy. In the past few decades, considerable developments in case of in vitro nucleic acid detection using small amounts of sample12 and in vivo intracellular measurements13 for diagnostic purposes have been reported. Important applications like gene expression profiling (e.g., for cancer detection),14 genotyping (e.g., for controlling infectious diseases)15 and bacterial detection (e.g., in monitoring genetically modified pathogenic bacteria)16 have also been made. For nucleic acid sensing, the target analyte is typically a stretch of nucleic acid sequence, be it ssDNA/ssRNA/dsDNA,17 or a nucleic acid motif, e.g., RNA G-quadruplex.18 In most cases, sensing is performed via complementary base-pair hybridization, so that detection is sequencespecific.17 In such cases, the sensor probe is a fully or partially complementary nucleic acid sequence, which is made available to the target for hybridization interactions to occur. Another type of nucleic acid sensing probe is nucleic acid aptamer, which is a single-stranded DNA/RNA sequence (or chemically modified DNA/RNA), usually 10-100 nt long, with specific threedimensional structures.19 Nucleic acid recognition by aptamers, for example, stem-loop probes, is also performed via complementary base-pair hybridization, where drastic conformational change of the aptamer probes upon hybridization is detected.20 So far, the majority of the investigations on nucleic acid analyses have been performed using DNA sensor probes. Although DNA-based sensors have found wide applications in microscale1 as well as nanoscale experiments,21-23 bioactivity of DNA probes may suffer on solid surface due to non-specific DNA-surface interactions (for example, gold-nitrogen interactions) through the

3 ACS Paragon Plus Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 45

exposed nucleobases. Moreover, the natural nucleic acid sequences are prone to nucleaseinduced enzymatic degradation.24 A library of synthetic alternative nucleic acid analogues like locked nucleic acid (LNA), peptide nucleic acid (PNA), 2´OMe RNA, threose nucleic acid (TNA), hexitol nucleic acid (HNA), arabinonucleic acid (ANA), fluoro-arabinonucleic acid (FANA) and glycol nucleic acid (GNA) [Figure 1] has been developed over the last two decades.2533

Amongst all these alternative nucleic acid probes, which are also called xeno nucleic acid or

XNA probes (where 'xeno' is for 'alien'), the LNA and PNA probes are the most promising probes for nucleic acid sensing. Both LNA and PNA are nuclease-resistant, can form thermally stable duplexes with DNA/RNA sequences, and sufficiently long sequences of LNA/PNAincorporated sensor strands can be generated.25,26 LNA is considered to be the best probe for SNP analysis in solution so far, while performance of the PNA probe has also been commendable. In this article, the essential steps that are required for nanoscale nucleic acid sensing and some recent successes using LNA and PNA probes will be presented.

Primary requirements for nucleic acid detection at nanoscale The basic steps in a nucleic acid sensing experiment are (a) sensor probe immobilization at solidliquid interface, (b) target approach and duplex formation upon complementary base-pair hybridization, and (c) monitoring hybridization (or dehybridization) event by means of a signal transduction method [Figure 2A].

Probe immobilization An obvious outcome of appropriately designed sensor probe immobilization is high probe density value, yet not too high that target entry inside the sensor layer is inhibited. Properly optimized sensor probe layer34 ensures nearly upright probe orientation that allows facile access


Page 5 of 45 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

of the sensor probes by target, and disfavor non-specific interactions between sensor probe and the underlying adjacent areas of substrate (if uncovered). For stable immobilization chemistry that keeps the probes surface-anchored yet functional over at least few weeks to months, careful consideration of ‘molecular adsorption’ and ‘molecular assembly' is essential. While some degree of molecular adsorption can occur at any solid-liquid interface, of greater interest are the systems where one or more components of liquid phase are strongly adsorbed at solid-liquid interface. This results in a higher adsorbate concentration at interface than in the bulk solution and a sizeable lowering of interfacial tension as described by Gibbs adsorption isotherm [Figure 2B]. Effects of such strong adsorption are of practical importance since it allows manipulation of solid-liquid interface to the maximum advantage. Here, an ideal site-filling of strongly surfaceactive solutes in terms of Langmuir adsorption isotherm [Figure 2C] demands that the solutes are non-interacting to each other. This is impracticable in case of the negatively charged oligonucleotide, unless suitable ionic control by buffer modulation is in place to shield the charge that is accumulated in the interfacial region upon oligonucleotide fixation. Another strategy could be to engineer the backbone to make it uncharged and/or conformationally rigid so that backbone orientation becomes more upright and therefore, be subjected to less electrostatic interactions with the adjacent oligonucleotides. Typical immobilization methods that result in strong molecular adsorption, include covalent binding, for example, 5′-amino-modified ssDNA onto aldehyde or epoxy-modified surface via formation of imine or secondary amine;35 noncovalent binding, for example, affinity binding as in case of avidin-biotin interactions;36 and chemisorption via covalent/near-covalent interactions, for example, adsorption of 5′-thiol modified ssDNA oligonucleotide onto gold surface via formation of gold-sulfur linkage37 [Figure 2D].


Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 45

Since formation of a well-organized and compact nucleic acid film onto a solid substrate depends on a delicate interplay of 'adsorbate-substrate' and 'adsorbate-adsorbate' interactions, molecular self-assembly interactions that originate from non-covalent links like hydrogen bonds, electrostatic interactions, hydrophobic interactions, and van der Waals interactions, need to be strong enough to overcome the intrinsically unfavorable entropy that is involved in bringing the surface-anchored molecules in aggregated state and drive the molecules into a low-energy stable state. This happens when the adsorbate concentration is within a certain range, below which the layer compactness and molecular ordering is lost, and above which a propensity for multilayer and/or disordered aggregate formation is observed.

Probe access and target recognition Since ssDNA, having a low persistence length value,38 can be conformationally flexible, the sensor strand may adopt a coiled conformation, making full access of it by target difficult. The nucleic acid sensing layers therefore perform relatively well when the sensor strands are of short length (10-15 mer), although detection of long target sequences (20 mer and above) can be difficult, unless a segment of the target sequence is of comparable length to that of sensor probe and complementarity of this segment to sensor probe is sufficient for target detection. In this context, it is tempting to propose that the LNA probe is more appealing compared to DNA probe as LNA probe offers conformationally less flexible backbone due to the locked sugar conformation. It can therefore be oriented more upright than DNA probe and be more accessible to target. So far, the nucleic acid detection reports have been made mostly on recognition of ssDNA, and rarely on dsDNA detection.39 The primary requirement for dsDNA detection is an ability of


Page 7 of 45 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

triplex formation that the sensor probe must possess. For example, PNA probe can either form PNA/DNA2 triplex or PNA/DNA2 duplex invasion which is dependent on PNA concentration, oligomer length and composition.40 In case of LNA-containing 'triplex forming oligonucleotides' (TFOs) that are conformationally pre-organized for major groove binding,41 it has been observed that reduced LNA content at the consecutive position of 3′-end destabilized the triplex structure, whereas the presence of Twisted Intercalating Nucleic Acid (TINA) at the 3′-end of TFO increased the rate and extent of triplex formation.41 It has also been shown that double-strand invasion (DSI) complex can be efficiently formed by LNA-containing oligomers.42 In case of nanoscale biosensors, effective designing of the biosensing surface is crucial for minimization of resources, so that manufacturing and usage cost can be reduced. Defects in sensing film and presence of impurities in sensing media can cause difficulty in achieving high sensitivity and target specificity, as recognition events that are detected here are less in number compared to that in ensemble experiments. Substrate cleanliness is important since effective molecular anchoring onto substrate surface and subsequent layer formation depends on availability of the anchoring sites. Preparation of a proper surface cleaning protocol and its effective use is therefore a pre-requisite for sensing film preparation and reliable bio-detection. Differentiating repeatable target-specific interaction signal from repeatable non-specific (or erroneous) signals is an issue that one may encounter in nanoscale/single molecule level detection. A judicious choice of the type of control experiment and careful interpretation of control response can be pivotal in this respect. When data acquisition is justified by control response and by the number of data sets (ideally large, especially in case of the single molecule level measurements), statistical analysis of the total data set is essential for obtaining an idea about the behaviour of a 'distribution' rather than 'an event'.


Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 45

Transduction methods Much of a biosensor's success depends on the suitability of transduction mode that is involved in sensor operation. A myriad of transduction modes, that can broadly be divided in two categories - label-inclusive and label-free, can be employed for nucleic acid sensing. For label-inclusive detection, methods based on fluorescence measurement,43 chemiluminescence measurement,44 and colorimetry45 are used. Label-free optical techniques, for example, nanomechanical detection of microcantilever bending via monitoring LASER deflection;21 detection by means of an optical spectrum using IR, UV or other types of radiation;46 a fiber optic-based approach,47 surfaceenhanced

Raman

spectroscopy

(SERS),48

surface

plasmon-based

detection,49

and

interferometry50 have been demonstrated. Non-optical label-free transduction modes include electrochemical, mechanical (piezoelectric/piezoresistive), acoustic, chemical, magnetic and thermal methods. The surface acoustic wave (SAW)-based device presents attractive option as it allows operation within wide dynamic range, is sensitive, reliable, and can be miniaturized.51 Magnetoelectronics has emerged as one of the novel technologies for detection of oligonucleotide hybridization by means of relaxivity change.52 Amongst all the transduction modes, electrochemical and mechanical transductions are the most commonly used, due to versatile applicability, high specificity and sensitivity that they usually offer. Mechanical transduction

in

case

of

(a)

microfabricated

cantilever

(LASER-based

or

piezoelectric/piezoresistive detection), either for surface stress sensing21,53 or mass sensing,54 and (b) oscillating piezoelectric crystal resonating at its natural resonance frequency, for example, quartz crystal microbalance55 offer encouraging possibilities for point-of-care applications. Microcantilever movement associated to bio-molecular complex dissociation can be another avenue where nanomechanical transduction is employed.7 This strategy of AFS-based single


Page 9 of 45 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

molecule force spectroscopy (SMFS) offers high spatial (nm), temporal (ms) and force (pN) resolution and the capacity of detection at molecular resolution. Being label-free, it can provide direct information on a system/event.

Thermodynamics and kinetics treatment Thermodynamic treatment in terms of solution phase enthalpy values (∆H), for example, in differentiating stabilities of DNA/DNA and RNA/RNA duplexes,56 has been reported, although not performed at nanoscale. Since nanoscale measurements may involve non-equilibrium events, for example, monitoring duplex dehybridization by force-induced rupture using SMFS approach, where movement of the dehybridized strand is restricted in a single direction, analysis of such data requires suitably modified thermodynamical/kinetic considerations. In order to extract equilibrium energy information from SMFS experiments, a Boltzmann-weighted work-averaging method57 based on Jarzinsky's equality between the thermodynamic free energy differences and the irreversible work done along non-equilibrium trajectories (see equation 1) can be applied

….............………………………………… (eq. 1)

where, ΔG is Gibb’s free energy, F(z) is measured force, the second term in the exponent is initial cantilever deflection, kB is Boltzmann constant and T is experimental temperature. One additional constraint in nucleic acid hybridization event at solid-liquid interface is that while the target strand approaches freely from aqueous phase, the other strand remains immobilized at a fixed point on the substrate. Suitable modifications in the adsorption isotherms are therefore needed so that biosensing data acquired from measurements at solid-liquid interface can be fitted


Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 45

and the kinetic rate constants can be obtained. For example, the concentration-dependent data obtained on label-free detection of target DNA sequences by ssPNA probes using a microcantilever array-based assay [Figure 3] was fitted to the modified Langmuir adsorption isotherm model, from which the thermodynamic equilibrium dissociation constant, K-1, was derived (equations 2 and 3). Surface coverage = Cmax·C/(K-1 + C) …………………………………………………....…. (eq. 2) where, surface coverage is ratio of the number of bound molecules to the total number of available binding sites, C is target concentration in solution and Cmax is saturation loading capacity.58 By assuming that the cantilever bending signal is proportional to surface coverage, and to the differential surface stress (Δσ) generated onto the cantilever, the following relation could be used to obtain the dissociation constant value Surface stress (Δσ) = A·Cmax·C(K-1 + C) ................................................................................ (eq. 3) where, A is a proportionality constant. At full saturation, it is a measure of the cantilever’s nanomechanical response, independent of target concentration. Importantly, the dissociation (or association) constant value as obtained from the microcantilever array-based detection was found to be of the same order as the value obtained in pure solution medium.58 Such a striking similarity indicates unchanged affinity to target hybridization at the interface, compared to solution phase, most likely due to a Langmuir-like probe coverage where the probes are sufficiently apart and non-interacting to each other. For non-ionic PNA probes such coverage is plausible, effectively offering a less challenging electrostatic environment to target DNA for its approach and binding, compared to the negatively charged DNA probes. Similar parity between the solution and the interfacial kinetics has not been commonly observed in case of DNA probes and reduced affinity to target hybridization has been reflected in the equilibrium constant values


Page 11 of 45 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

irrespective of the nucleobase sequences. Since the above-mentioned solution vs. interface comparative treatment is crucial as the most biologically relevant phase is 'solution' and any nanoscale experiment that employs a solid surface deviates from this ideality, careful choice of sensor probe backbone and/or ionic composition of hybridization buffer is important for the target to overcome the electrostatic barrier it faces at solid-liquid interface.

When natural is not enough: The XNA world The natural nucleic acids, DNA and RNA, are essentially periodic polymeric chains composed of nucleoside building blocks, where each two adjacent nucleosides are linked by a phosphodiester bond. Although both the nucleoside and the phosphodiester linkage contribute to the physicochemical properties of the nucleic acid polymer, there is a dominant contribution by the polyanionic phosphodiester backbone,59 which decouples the physicochemical properties of a nucleic acid from its information content (i.e., its nucleobase sequence). Thus all the nucleic acid sequences display broadly similar physicochemical properties in sharp contrast to proteins. In case of the latter, a single mutation in the amino acid sequence can lead to a radical change in the physicochemical properties, for example, solubility. In case of XNA probes, the physicochemical properties can be varied as per requirement, since XNA synthesis involves replacement of the sugar-phosphate backbone of DNA/RNA by a synthetic backbone (non-sugar and without phosphodiester linkage). The XNA may also consist of modified sugar units, connected to each other by phosphodiester links,25-33 or modified nucleobases as present in 5-aminoallyl-dUTP, 8histaminyl-dATP, 5-guanidinoallyl-dUTP, 5-aminoallyl-dCTP, and 8-histaminyl-dATP.60 Amongst all the XNA types, LNA and PNA are the most widely studied in solution phase due to efficient hybridization capacity that they exhibit. GNA probe cannot form stable antiparallel


Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 45

cross-pairs with DNA,32 and ANA can only weakly associate to ssDNA target.30 Although HNA retains the ability to pair with DNA/RNA, selective incorporation of more than two anhydrohexitol units has been found to be difficult.33 The probes ANA and F-ANA are reported to induce the widespread RNaseH cleavage when forming duplex with RNA.31 Although TNA can form thermally stable duplexes complementary to DNA/RNA sequences and is nucleaseresistant,28 its solution phase hybridization behaviour has not been as extensively studied as LNA and PNA probes. In our studies, we therefore explored the capacity of LNA and PNA probes as promising alternatives to DNA probe for nanoscale nucleic acid sensing.

The LNA probe and its application LNA is a polymer of 2′-O, 4′-C-methylene-linked β-d-ribonucleotide monomers [Figure 1], where the linkage of 2′-O and 4′-C atoms via methylene bridge restricts the 3′-endo conformation in a RNA mimicking N-type sugar conformation.61 Such locking of the furanose ring imparts higher stability to the duplex formed between LNA nucleotide and target DNA/RNA. LNA probe is capable of binding with complementary DNA/RNA obeying Watson-Crick base pairing rule with higher affinity and sequence specificity compared to DNA probe. This is reflected in higher Tm value of the LNA-DNA/LNA-RNA duplex compared to that for the DNADNA/DNA-RNA duplex.62 LNA is water-soluble and can have multiple water bridges that provide it with extra stability compared to DNA/RNA.63 Since LNA is a conformationally restricted molecule, its higher structural rigidity can prevent non-specific interactions with the solid substrate, and support an upright backbone orientation,64 which is conducive to effective recognition of the target sequences. These advantages of LNA, i.e., enhanced thermal stability of the LNA-containing duplexes, water solubility, and oriented disposition, combined to the


Page 13 of 45 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

nuclease-resistant nature of LNA,65 indicate useful applications of the LNA probes both for in vivo and in vitro diagnostics. LNA phosphoramidites (available with all the four nucleobases A, C, G and T) and their oligomers are commercially available, and LNA nucleotides can be mixed with those of the natural nucleic acids for generating heterogeneous chimeric probe molecules called mixmers. This makes LNA a flexible tool in nucleic acid diagnostics and nucleic acidbased therapeutics.66 It has been shown that the highest Tm increase per LNA modification and the best mismatch discrimination are achieved when short LNA sequences (~10 bases) are used.67 The capacity of the LNA probes has been exemplified in capturing PCR amplicons in a solidphase hybridization assay,68 in microarrays for detection of multiple miRNAs,69 and in SNP scoring using short LNAs by means of fluorescence polarization detection.70 Dually labeled LNA hairpin probe with biotin (for streptavidin-based immobilization) and a carboxyfluorescein (FAM) molecule (as an affinity tag for HRP) has been used for SNP detection.71 The effect of incorporation of LNA nucleotides into RNA aptamer specific to the HIV-1 TAR RNA element has also been reported.72 Fang et. al. reported LNA microarrays for detection of LNA-captured miRNAs by first addition of poly(A) tail to the surface-bound miRNA via poly(A) polymerase reaction, and then hybridization with T30 DNA-modified AuNPs for AuNP-amplified surface plasmon resonance imaging (SPRI) detection.69 LNA has been incorporated in molecular beacons (MB), resulting in combined advantage of sensitivity that MB-based detection imparts and the high binding affinity and stability that the LNA-containing nucleic acid sequences exhibit. Wang et. al. engineered MB with LNA backbone to generate novel probes that provide high thermostability, considerable affinity towards complementary sequence, high selectivity for detection of mutants, nuclease resistance, and overall reduction of false positive signals.73 LNA-


Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 45

modified capture probe (18-mer) has been used for detection of chronic myelogenous leukaemia via hybridization with the BCR/ABL fusion gene using differential pulse voltammetry, where the detection limit was found to be 9.4×10−13 mol·L−1.74

Use of LNA probe in nanoscale studies In spite of the earlier successes of LNA-incorporated probes, reports on nanoscale nucleic acid detection are limited in number.75 An ability of target recognition at molecular resolution that could assist in nanoscale detection of few molecules is therefore an aim worth pursuing. We developed simple ways of generating self-assembled, ordered LNA monolayers onto silicon and gold(111) surfaces.7,64 Stronger DNA recognition signal (4-4.5 times) was observed using LNA probe, compared to its DNA counterpart, as found from fluorescence spectroscopic measurements [Figure 4A]. Importantly, the surface-anchored LNA probe could differentiate between a fully complementary DNA target sequence and a singly mismatched sequence, where the mismatch discrimination ratio was approximately twice greater compared that observed in case of DNA-based detection [Figure 4A].64 From the molecular orientation sensitive reflection absorption infrared (RAIR) spectroscopic traces, it was found that the LNA backbone orientation was more upright than DNA backbone as considerably more resolved spectra and more intense peaks at the frequency values corresponding to RAIR-active modes were observed in case of LNA, compared to DNA [Figure 4B and C]. This advantageous disposition of the LNA backbone could be one of the reasons behind its superior performance as a sensor probe, than DNA probe, since a more upright backbone means less non-specific interactions between the sensor probe and substrate, and more accessibility of the sensor strands by target. The orientational advantage of LNA over DNA could be related to the conformationally more rigid


Page 15 of 45 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

LNA backbone due to bridged (or locked) sugar structure. That conformational rigidity of LNA was operative in case of the surface-anchored LNA probes was revealed from the SMFS study on LNA-DNA duplex unbinding events.7 It is important to note here that conformational rigidity of LNA backbone is in favour of a Langmuir-like non-interacting probe coverage although such coverage would have been unlikely due to the negatively charged backbone of LNA. Therefore, as a further measure, we tuned ionic strength and pH of immobilization buffer so that an optimal range of LNA probe density as required for maximizing single base mismatch discrimination capacity could be identified [Figure 5].34 Counter cation condensation around the interfacial LNA structures meant shielding of LNA backbone charge that would maximize LNA probe density. However, since effective target entry requires sufficient space around each sensor probe, an increase in the probe coverage beyond the optimal probe density window led to a drastic reduction in target recognition efficiency as indicated by fluorescence intensity measurements [Figure 5]. The probe density was found to be greater in case of LNA compared to DNA at similar immobilization condition.34 Since LNA backbone is conformationally more rigid than the DNA backbone, packing a greater number of sensor stand within a certain volume at the solidliquid interface is more likely in case of LNA probes, although both DNA and LNA are negatively charged. Interestingly, introduction of spacer molecules within the LNA layer to form a mixed monolayer did not influence LNA backbone orientation [Figure 5], which is a major shift from DNA-based sensing as spacers are found to play a beneficial role therein. We propose that this exercise of finding an optimal probe density by systematic designing in view of structural/functional aspects of molecular organization at solid-liquid interface can be crucial for nanoscale/molecularly resolved nucleic acid detection.


Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 45

In order to obtain a quantitative idea on the performance of LNA layer, we developed an analytical method for measurement of melting temperature (Tm) of the surface-anchored nucleic acid duplexes.76 In this method, fluorescence intensity of the dehybridized target sequences (Cy3-labeled) can be monitored at different temperatures (for example, a range of 25 ºC to 70 ºC or to a temperature upper limit as indicated by complete dehybridization over the total film area). Upon plotting the intensity values against temperature, a sigmoid melting curve can be obtained and from the mid-point of this curve, the on-surface Tm value can be recorded [Figure 6]. A quantitative assessment of the effectiveness of any type of nucleic acid probe in a wide range of hybridization environment is therefore possible using this approach. Estimation of the on-surface Tm of nucleic acid duplex is necessary to identify the temperature range above which complete duplex denaturation can be ensured. Optimization of hybridization temperature is also much dependent on Tm value as typical hybridization temperature recommended for nucleic acid hybridization should be at least 20 °C below than the calculated Tm of the duplex to achieve a near-maximum hybridization rate.77 Particularly, in the case of LNA probes (different lengths), typical hybridization temperature recommended should be between at least 15 °C - 30 °C below than the Tm value.78 While mismatch discrimination capacity of both DNA and LNA probes can be improved at the solid-liquid interface in comparison to solution phase, LNA probes excelled DNA probes in performance both in solution phase and when confined within the interfacial space.76 We found it useful to apply different types of cations in hybridization buffer in order to maximize the onsurface LNA-DNA hybridization signal. Use of bivalent Mg2+ was found to be more effective in single base mismatch discrimination compared to monovalent sodium, trivalent spermidine and tetravalent spermine.76 In case of on-surface detection, we found that although rising salt


Page 17 of 45 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

concentration increased Tm values of the fully matched duplexes, the influence of salt concentration on singly mismatched duplexes was nearly negligible [Figure 6]. Single base mismatch discrimination ability of the LNA sensing layer could thus be effectively enhanced with increasing salt concentration. In our case, a noteworthy application of the optimized LNA film structure could be realized in attaining a high percentage (~80%) of molecule-by-molecule detection of target DNA sequences, using AFS-based single molecule force spectroscopy (SMFS) approach. Such an experimental development means that detection of even few strands/molecules, as generally required for nanoscale sensing, is possible using LNA probes. In SMFS, target is detected by monitoring force-induced unbinding (or induced dehybridization) event that takes place during cantilever retraction stage [Figure 7A].7 Since nucleic acid hybridization interactions generate stabilizing contacts between the participating strands, where weak forces like hydrogen bonding, hydrophobic and electrostatic interactions drive the hybridization process and lead to an energy reduction, a definite amount of pulling force (unbinding force) needs to be applied to the cantilever, for duplex dissociation to take place [Figure 7A]. At single molecule level, as recognition process goes through a diverse conformational combination of the single molecule binding pair, a statistical distribution of the unbinding force values is obtained from recognition events [Figure 7B]. Upon careful statistical analysis of the data sets, a meaningful unbinding force value can be obtained. A major advantage of the SMFS-based detection turns out that nucleic acid amplification is not required as the approach is adept at molecular level detection and requires a small amount (~10-20 nM) of target sample. Amongst other advantages, each event of target detection could be monitored at the millisecond time scale (every hybridization/dehybridization event taking place within 100 ms), which correlates well with the


Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 45

recently reported single molecule level DNA hybridization kinetics data, which is obtained using a 8 bp duplex in solution phase by a time-resolved, 3-D single molecule tracking (3D-SMT) approach.79 In the 3D-SMT study, a combined confocal-feedback 3D-SMT with time-domain fluorescence lifetime measurement permitted tracking of a freely diffusing (at speed 4.8 µm2s−1) ssDNA molecule in solution for hundreds of milliseconds and monitor multiple hybridizationdehybridization events occurring on the same molecule. We pursued SMFS for detection of dehybridization events with the understanding that at single molecule level, dsDNA unbinding is caused by thermal fluctuations and not by mechanical instability, since for thermal lifetime of the bond, which is too short compared to the rate of force loading, no unbinding event can be observed.80 Effective unbinding forces would depend on the rate of force loading and a functional relationship between bond lifetime and applied force. Schumakovitch et. al. extracted thermal lifetime from unbinding force measurements for the temperature range 6°C to 36°C that can be compared with lifetime measurements using the temperature jump (TJ) method near melting transition (35°C to 50°C) of the DNA duplex.81 They calculated activation enthalpy for DNA duplex dissociation, ΔHAFM = 300±42 kJ/mol, which is comparable with ΔHTJ = 375±16 kJ/mol. A reasonable agreement of bond lifetime from force measurements with the TJ values ensures that the Bell's ansatz correctly describes the dissociation kinetics under an applied force for the DNA duplex.82 From SMFS-derived unbinding force values, it was found that the surface-confined fully matched LNA-DNA duplex was nearly two times more stable than the fully matched DNA-DNA duplex.7 In order to understand the dissociation mechanism of the surface-anchored LNA-DNA duplex, we investigated dependence of the unbinding force values on force loading rate and fitted experimental data with the Bell-Evans model80,82


Page 19 of 45 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

F = (kBT/xβ) ln(rxβ/koffkBT) ..................................................................................................... (eq. 4)

where, F is unbinding force, kB is Boltzmann constant, r is loading rate, T is absolute temperature, xβ is length scale of the potential barrier on dissociation pathway and koff is kinetic off-rate constant for dissociation (at zero force). A Koff value indicative of low affinity, dynamic interactions, and a low xβ value, reflecting lesser conformational variability of the surfaceconfined LNA-DNA duplex compared to dsDNA, were obtained.7 The conformational rigidity of LNA means lesser reduction in entropy during duplex formation, compared to the more flexible DNA that undergoes a large degree of conformational change during binding/unbinding. The competition between entropic and enthalpic contributions is weaker in case of LNA probe that works in favour of the LNA-based detection system. We further investigated if the silicon-anchored LNA probe could differentiate amongst different types of nucleobase mismatches [Figure 7C] since for development of a diagnostic system for genetic screening, identifying different mismatch types is essential. It was found that any deviation from the most commonly formed pairs of nucleic bases A-T and G-C, could be identified and differentiated from each other in most cases.83 This assay was further tested for detection of the mutant gene sequences as relevant to the multiple drug-resistant tuberculosis (MDR-TB) bacteria [Figure 8].83,84 The most prevalent codon mutations leading to drugresistance of Mycobacterium tuberculosis could be detected in case of first-line drugs rifampicin and isoniazid.83,84 In this study, the unbinding force information could be translated into mapping of potential energy landscape with molecular resolution using Jarzinsky's equality treatment.57 The LNA-based SMFS assay that can detect any target sequence, provided it is not too long (

Nanoscale Nucleic Acid Recognition at the SolidâLiquid Interface

Recommend Documents

Nanoscale Nucleic Acid Recognition at the SolidâLiquid Interface