Impact of point-mutations on the hybridization affinity of surface-bound DNA/DNA and RNA/DNA oligonucleotide-duplexes: Comparison of single base mismatches and base bulges
© Naiser et al; licensee BioMed Central Ltd. 2008
Received: 10 January 2008
Accepted: 13 May 2008
Published: 13 May 2008
The high binding specificity of short 10 to 30 mer oligonucleotide probes enables single base mismatch (MM) discrimination and thus provides the basis for genotyping and resequencing microarray applications. Recent experiments indicate that the underlying principles governing DNA microarray hybridization – and in particular MM discrimination – are not completely understood. Microarrays usually address complex mixtures of DNA targets. In order to reduce the level of complexity and to study the problem of surface-based hybridization with point defects in more detail, we performed array based hybridization experiments in well controlled and simple situations.
We performed microarray hybridization experiments with short 16 to 40 mer target and probe lengths (in situations without competitive hybridization) in order to systematically investigate the impact of point-mutations – varying defect type and position – on the oligonucleotide duplex binding affinity. The influence of single base bulges and single base MMs depends predominantly on position – it is largest in the middle of the strand. The position-dependent influence of base bulges is very similar to that of single base MMs, however certain bulges give rise to an unexpectedly high binding affinity. Besides the defect (MM or bulge) type, which is the second contribution in importance to hybridization affinity, there is also a sequence dependence, which extends beyond the defect next-neighbor and which is difficult to quantify. Direct comparison between binding affinities of DNA/DNA and RNA/DNA duplexes shows, that RNA/DNA purine-purine MMs are more discriminating than corresponding DNA/DNA MMs. In DNA/DNA MM discrimination the affected base pair (C·G vs. A·T) is the pertinent parameter. We attribute these differences to the different structures of the duplexes (A vs. B form).
We have shown that DNA microarrays can resolve even subtle changes in hybridization affinity for simple target mixtures. We have further shown that the impact of point defects on oligonucleotide stability can be broken down to a hierarchy of effects. In order to explain our observations we propose DNA molecular dynamics – in form of zipping of the oligonucleotide duplex – to play an important role.
DNA microarray technology relies on the highly specific binding affinity of surface-tethered DNA probe sequences to complementary target sequences. Nucleic acid hybridization, the sequential base pairing between complementary probe and target strands, results in the formation of stable double-stranded duplexes. In microarray hybridization assays single-stranded nucleic acid targets – contained in a complex mixture of diffierent target sequences in solution – freely diffuse over the surface-tethered probes until they are captured by a complementary probe. Target strands often carry fluorescent dye labels to enable quantitative detection of the individual target species. Hybridized targets can be identified by the position of the corresponding microarray features (each containing one particular species of surface-tethered probe strands) within the regular grid of the DNA microarray.
In DNA microarray applications, along with a high binding affinity (providing sensitivity), a high specificity of probe-target hybridization is required to discriminate between sometimes very similar homologous sequences. Binding specificity is particularly important in genotyping applications where Single Nucleotide Polymorphisms (SNPs), genetic variations of single bases, are concerned. SNPs determine genetic individuality, but also predisposition to a variety of genetic diseases, response to drugs, pathogens, chemicals and other agents. SNPs are of great interest not only for genetic research but also for medical diagnostics and therapy [1, 2].
SNPs and point-mutations can be detected by means of relatively short 10 to 30 mer probes: Already a single mismatched (MM) base pair can result in a significant decrease of the duplex binding affinity with respect to the corresponding perfect matching (PM) duplex .
The binding affinity of mismatched duplexes – in bulk solution – is commonly predicted on the basis of the nearest-neighbor model [4–6]. A recent study by Pozhitkov et al.  revealed a poor correlation between predicted duplex binding affinities and actual hybridization signal intensities implying that the thermodynamic properties of oligonucleotide hybridization on DNA microarrays are by far not understood. In DNA microarray experiments the binding affinity of mismatched oligonucleotide duplexes is governed not just by nearest-neighbor parameters – as in solution-phase hybridization – but mainly by the position of the defect [7–10]. Furthermore, the secondary structure of the long target strands  and various surface effects  have a significant influence on the microarray binding affinity.
Fluorescently labeled target oligonucleotides used in this study.
Target sequence (5'→3')
After the current article was submitted, we became aware of further related studies in this area. Suzuki et al.  performed hybridization on custom NimbleExpress™ arrays (Affymetrix Inc.) to investigate the influence of the probe length and mismatch position on single base MM discrimination. Fish et al.  performed a direct comparison between hybridization signals (perfectly matching and mismatched duplexes) from spotted microarrays and measured thermodynamic melting parameters (determined by differential scanning calorimetry in bulk solution). They report a linear relation between the duplex free energy and the microarray hybridization intensity.
The focus of the present paper is on the impact of various defect types (single base mismatches and single base bulges) on the hybridization signal.
Our microarray hybridization experiments performed in this study provide quantitative information on the binding affinity of individual mismatched duplexes by means of the hybridization signal intensity (fluorescence of hybridized targets). Since the absolute hybridization signal intensities of the different sequence motifs employed in this study (Tab. 1) are subject to a large variation (often larger than between mismatched and corresponding PM hybridization signals) we compare the MM hybridization signals with the corresponding PM hybridization signals. Our experiments – experimental details (probe sets, hybridization signal normalization etc.) are explained in the Methods section – provide a measure for the mismatch discrimination with respect to the corresponding PM binding affinity, rather than an absolute measure for the MM binding affinity. The discrimination between the hybridization affinity of point-mutated probes and corresponding perfect matching probes depends on the stability of the particular probe sequence. In agreement with  we observed that the (more stable) 25 mer probes are less discriminative with respect to point defects than the shorter 16 mer probes. Discrimination is also reduced for sequence motifs stabilized by a higher CG-content.
MM defect position and hybridization affinity
Influence of the mismatch type in DNA/DNA duplexes
In the following we use the notation of the mismatch base pair X·Y consisting of the mismatched base X in the probe sequence and the base Y in the target sequence. To investigate how the particular MM-types X·Y affect duplex stability we measured probe-target-affinities for 25 different sequence motifs. Microarray hybridization experiments with single base mismatch probe sets as well as the extraction of their hybridization signals, which reflect duplex stability, are described in more detail in the Methods section. Owing to the limited number of available target oligonucleotides we restricted base substitutions to the probe sequences. The PM hybridization signals of the different 16 mer sequence motifs display a strong variation (up to a factor 20). Since the relative hybridization signal intensities within the individual probe sets are largely unaffected by this variation, we normalize the "defect profiles" by division with their standard deviation. The resulting database comprising normalized hybridization signals from about 1000 different single MM probe sequences, enables categorization of the binding affinities according to the mismatch type.
We compared the MM-type related hybridization signal deviations δI mp from the mean MM profiles (Fig. 2B) to predicted Gibbs free energy differences between MM and corresponding PM duplexes. δ were determined from mismatch nearest-neighbor thermodynamic parameters . Our analysis (shown in Additional file 2) indicates a decreasing trend of the δI mp values with increasing δ. Moreover, we observed that single base mismatches with two A·T flanking base pairs tend to provide a better mismatch discrimination than mismatches flanked by two C·G base pairs.
DNA/DNA single base bulge defects
Single base insertions and deletions owing to a surplus unpaired base in one of the two strands result in bulged duplexes. In our experiments (sequence data and hybridization signal raw data is provided in Additional file 3) the bulged base is located on the surface-bound probe strand, whereas in duplexes with single base deletions (on the probe sequence) the bulge is on the target strand.
Interestingly, systematically increased hybridization signals (with respect to the averaged hybridization signal level from other defect types at the same position) have also been observed for certain Group I bulges: For G-insertions next to a T (e.g. in Fig. 5 at base position 15) we frequently find increased binding affinities similar to that of Group II bulges.
We further analyzed the degree of correlation between the binding affinities of probes with different insertion bases X and Y (see Additional file 5): A clear correlation appears between the hybridization signals of probes with T- and G-insertions, and also, though less distinct, between A- and C-insertions. In contrast, we observed an anti-correlation between G- and A-insertions.
DNA/DNA versus DNA/RNA mismatch and bulged hybridization
For bulged duplexes we did not observe significant defect-type specific differences between RNA/DNA and DNA/DNA hybridization. The hybridization signal and sequence data from the microarray hybridization experiment are provided in Additional file 8.
Single base insertion, deletion and mismatch defects in comparison
Defect profiles for MMs and base bulges (Fig. 6) exhibit a very similar quantitative influence from defect position in DNA/DNA as well as in DNA/RNA complexes. For individual sequences the mean trough-shaped profile can be altered: Fig. 3 shows deformations of the trough-like profile on scales much larger than the size of a base pair.
Single base MM discrimination also depends on the type of MM base pair and the corresponding PM base pair (which has been substituted by the MM). Hybridization signals of MMs (normalized with the respect to the PM hybridization signal) originating from C·G base pairs are about 25% smaller (in the median) than for MMs from A·T base pairs. Single base deletions affecting C·G base pairs result in about 30% smaller hybridization signals than deletions affecting A·T base pairs. The deletion profile in Fig. 6 (orange dashed line) shows that the local ups and downs of the profile curve correlate with deletions affecting either A·T or C·G base pairs. Thus, for MMs and single base deletions it is the type of base pair affected by the point-mutation, which determines the impact on hybridization affinity to an important degree, however, it is still less important than defect-position.
We also observe a noticeable influence of the next-neighbor bases of the mismatch (see Additional file 2).
Dominating influence of defect position
We observe that defects located in the center of the oligonucleotide duplexes are significantly more destabilizing than defects at the ends . Similar influence of the MM position has been reported previously from other microarray based studies [7, 9], and also – although sparsely – from hybridization experiments in solution [25, 26]. The limited data in solution may be due to the technical difficulty of studying a large number of different probes. Quantitatively, in accordance with  we have identified MM position (relative to the duplex ends) as the strongest influential factor on the hybridization signal, when compared to MM-type and nearest neighbor effects.
The well-established two-state nearest-neighbor model, which has proved to be reliable for the prediction of duplex stabilities in solution-phase, does not regard the position of the (mismatched) NN pairs . We propose that a model for the prediction of microarray binding affinities should also include the position of the NN pairs – in particular in case of mismatched NN pairs. Affinity models for microarray hybridization considering a positional dependence of the nearest-neighbor parameters have been previously discussed in [12, 27–30].
We observe a very similar position dependence for single base bulge defects as for single base mismatches. Also, the magnitudes of the impacts of the MMs and base bulges on the hybridization signal are very similar (apart from the relative high binding affinity of Group II bulges). This consistency suggests a common origin of the positional influence, independent of defect type.
Sterical crowding at the surface, as suggested by Peterson et al. , can in principle reduce the accessibility of the probe surface-bound 3'-ends and can thus decrease the impact of defects located near this end. However, in our case we observe largely symmetrical intensity profiles with respect to both ends of the probes (Fig. 2).
Focusing on individual probe sequence motifs we observe, that the positional influence does not only depend on the defect-to-end distance, but also has a sequence-dependent contribution. This indicates that the impact of a defect also depends on the stability of the local sequence environment (beyond the nearest neighbors). Since there are no long range molecular forces, we infer that the molecular dynamics must play a role, effects like breathing bubbles or zipping could be at the origin. This influence of the duplex sequence and the observed symmetry of the defect positional influence with respect to the duplex ends suggest that end-domain opening (i.e. sequential unzipping of the double-helix from the duplex ends) must be suspected to be a key mechanism for understanding the influence of defect position on duplex stability.
Influence of the MM-type
Removing the positional influence in our data, we see that single-base MMs introduced at the site of a C·G base pair result in a larger decrease of the hybridization signal (with respect to the PM hybridization signal) than MM defects affecting A·T base pairs. The same applies for single base deletions (see Fig. 6). These experimental results (Fig. 4), in accordance with nearest-neighbor thermodynamic parameters for Watson-Crick base pairs , mainly reflect the increased base stacking and hydrogen bonding interactions of C·G base pairs. We observe a positive correlation between the experimentally determined single base mismatch discrimination and predicted free energy increments δ (between MM and PM duplexes) on the basis of the nearest-neighbor model – for details see Additional file 2. A similar correlation (between log2(PM/MM) hybridization signal values and δ) has been reported previously in .
We emphasize the good correlation between our DNA/DNA MM stability order (Fig. 12e) and the corresponding results of Wick et al.  (the MM stability order in Fig. 12d was extracted from the plot of log2(PM/MM) hybridization signal values in Fig. 5a in ). A major difference, however, occurs for the MM-pair G·G, which is the least stable in our study. Wick (and also Sugimoto ) found G·G to be one of the most stable MMs. Interestingly, however, Pozhitkov et al.  – in accordance with our results -identified G·G as one of the least stable MM-types.
Our direct comparison between DNA/DNA and RNA/DNA hybridization on microarrays reveals – for RNA/DNA duplexes – an increased destabilization of purine-purine mismatches, with respect to other MM types. An explanatory approach for the observed differences between DNA/DNA and RNA/DNA binding affinities is, that purine-purine MMs cause larger steric hindrance in the A-form hybrid duplexes than in the B-form DNA/DNA duplexes.
In contrast to  we did not observe that purine-purine mismatches in RNA/DNA duplexes are, in absolute terms, more discriminative than other MM-types.
Increased stability of Group II single base bulges
We observe significantly increased hybridization signals of single- base insertion defects in which the insertion base is placed next to a like-base. Our investigation shows that (on the microarray) the difference between Group I and Group II binding affinities δI bulge (inferred from the hybridization signal I) is distinctly larger than the defect-type related variation of binding affinities δI MM (see Fig. 7). For comparison, the free energy differences among the MM trinucleotide duplexes and (mismatched bases x and y; neighboring bases a and b unchanged; overline denotes complementary bases) span the range = 0.5 to 2.6 kcal/mol (calculated with MM nearest-neighbor free energies  for T = 37°C).
The increased stability of Group II bulges in comparison with Group I bulges has been investigated previously in solution rather than on microarrays [24, 34, 35]. According to Ke and Wartell  the increased stability of Group II bulges originates from positional degeneracy of the base bulge. Additional conformational freedom, entailing higher entropy, results in lowered duplex free energy (thus in increased stability). According to Zhu et al.  position degeneracy accounts for an average stabilization of -0.3 to -0.4 kcal/mol (in agreement with the theoretical estimate  of -R·T· ln 2 = -0.43 kcal/mol at 37°C) for a two-position degeneracy. Znosko et al.  reported Group II duplexes to be on average δΔG37 = -0.8 kcal/mol more stable than Group I duplexes. The latter value matches better our observation of Group II hybridization close to the perfect match hybridization signal.
Previous studies – including RNA/DNA hybridization
Tautz and coworkers  performed a mismatch study with 20 mer oligonucleotide microarrays fabricated by light-directed in situ synthesis with the Geniom® One instrument (febit biomed GmbH, Heidelberg). Similar as in our study, they compared normalized hybridization signal intensities.
However, an important difference between the experiments described in  and our experiments is the use of in vitro transcribed RNA targets  originating from ribosomal RNA. They observe a more pronounced destabilization by purine-purine MMs compared to our results.
A further study on the impact of MM stabilities in RNA/DNA duplexes, in solution rather than on a microarray surface, has been published by Sugimoto et al. . As discussed in  the destabilizing effect of purine-purine MMs is not observed by Sugimoto et al. . However, the stability order in , referring to ΔG37 values of mismatched trinucleotide duplexes, is considering absolute stability parameters, whereas [7, 9] and our study consider mismatch discrimination with the corresponding PM binding affinity as a reference level. Therefore, the comparability with the RNA/DNA stability order in  is limited. A recent work on the impact of single base MMs in RNA-interference (RNAi) – allele-specific gene silencing experiments  – is interesting in the context of this study, since here the sequence recognition is based on base-pairing between the guide strand (a single RNA strand which is bound to the RISC complex) and a complementary mRNA. Schwarz et al. (see Schwarz: table 5b) have shown that among all MM-types incorporated at position 10 of the guide strand (apart from the point mutations the sequence of the guide strand was preserved) purine-purine MMs resulted in the least silencing of gene activity (owing to a small binding affinity of the mismatched sequences), whereas U·G, C·U and U·U mismatches resulted in a very efficient gene silencing (see Fig. 12c). It is assumed that purine-purine MMs strongly interfere with the formation of an A-form helix between the guide strand and the target mRNA . This appears to be in accordance with the findings of Pozhitkov et al. on RNA/DNA MM discrimination. However, the inferred RNA/RNA mismatch stability order (shown in Fig. 12c) is not normalized with the corresponding PM stabilities, but rather reflects the absolute impact of the MM base pairs in a given duplex sequence and cannot be easily compared to our study and to .
We performed a comprehensive, array-based study on the influence of point defects on the binding affinity of oligonucleotide duplexes. Contrary to previous studies by others, we have employed well-defined hybridization conditions by using short, end-labeled oligonucleotide target sequences (one at a time to minimize competitive effects) and can therefore exclude that target secondary structure, steric hindrance, labeling or competitive effects are relevant for an explanation of the observed results.
In our microarray-based hybridization assays the binding affinity of mispaired duplexes is dominated by the influence of defect position. The influence of the defect-type is about half in magnitude, when compared to defect-position.
There is also an influence of the neighboring sequence, which has farther reach than the defect next neighbor. Although this long reach interaction must somehow be related to the base stacking energies, we did not find a simple description. We attribute so far unexplained long range effects, in particular a trough-shaped position dependence, to molecular dynamics. We propose a molecular zipping mechanism as a suitable explanation. Zipping agrees well with the observation that Group II bulges (bulges next to identical bases) have stronger hybridization signals than expected from previous data. Experimentally, it is not completely clear, whether the strong positional influence on oligonucleotide binding affinity is restricted to surface-hybridization or if it is also relevant for solution-phase hybridization (maybe to a smaller extend). The comparison to other related work [2, 7, 32], however, shows significant differences in the MM-type dependence of duplex binding affinities. Our comparative analysis of the impact of point defects on the binding affinity of DNA/DNA and RNA/DNA duplexes reveals that purine-purine MMs are more destabilizing in the latter. This may explain some discrepancies in the literature.
The use of DNA microarrays enables a detailed investigation of oligonucleotide duplex binding affinities producing a wealth of data in simple experiments. We demonstrate that important aspects (defect position influence, differences between DNA/DNA, RNA/DNA and RNA/RNA hybridization, surface and bulk hybridization) about the impact of point defects on oligonucleotide duplex binding affinities are not yet understood. Our results from simple, controlled experiments agree well with results from extracting data from complex DNA target mixtures [7, 9]. This shows that DNA hybridization on surfaces can be reproducible and quantitatively significant. Deviations from the behavior, which we describe here, are observed in microarray experiments and they must be due to complexity of DNA target mixtures.
All reagents were used as purchased without further purification. Unless specified otherwise aqueous solutions were prepared with nuclease-free Milli-Q water (18.2 MΩ cm).
Reagents used in dendrimer-functionalized substrate preparation
20 mm round cover glasses (Menzel-Gläser, Braunschweig, Germany); Deconex 11 UNIVERSAL (Borer Chemie AG, Zuchwil, Switzerland); (3-aminopropyl)-triethoxysilane (APTES) (Sigma-Aldrich); ethanol analytical grade (VWR, Germany); 1,2-dichloroethane (Cat. No. 6837.1, Carl Roth GmbH, Germany); phosphorous dendrimers with aldehyde moieties cyclotriphosphazene- PMMH-96 (Cat. No. 552097, Aldrich); potassium hydroxide (Carl Roth GmbH); sodium borohydride (99.99 %, Sigma-Aldrich).
Reagents and solutions used in light-directed DNA- Chip synthesis
RayDite™ photolabile 3'-nitrophenylpropyloxycarbonyl (NPPOC)-phosphoramidites (NPPOC-dA(tac), NPPOC-dC(ib), NPPOC-dG (ipac), NPPOC-dT) were purchased from Sigma-Proligo (Hamburg, Germany). Acetonitrile (ROTISOLV for DNA synthesis, water < 10 ppm, Carl Roth GmbH, Germany); Activator 42 0.25 M (Sigma-Proligo); iodine based oxidizer (part. no 401732, Applied Biosystems). Photo-deprotection is carried out in a mildly basic solution of 25 mM piperidine (99%, Aldrich) in anhydrous acetonitrile. Final base deprotection is performed in a 1:1 mixture of etylenediamine (analytical grade, Fluka) and ethanol (analytical grade, VWR, Germany). UV glue (Norland optical adhesive 60, Edmund optics) is employed to glue the chip after synthesis onto a stainless steel support.
The hybridization buffer comprises 5 × SSPE pH 7.4, with either 0.1% SDS or 0.01% Tween 20; the initial target concentration in the hybridization solution was 1 nM in all experiments.
Cy3-labeled target oligonucleotides (DNA and RNA) – see Tab. 1 – were synthesized by MWG Biotech AG (Ebersberg, Germany) and by IBA Nucleic Acids Synthesis (Göttingen, Germany).
Preparation of the phosphorus dendrimer-functionalized substrates
Dendrimer-functionalized substrates were prepared according to LeBerre et al. . For compatibility with the in situ synthesis process (coupling of phosphoramidite building blocks) the aldehyde moieties of the dendrimers are reduced to hydroxyl groups. Reduction is performed in an aqueous solution of 0.35% sodium borohydride (for 3 hours at room temperature, under gentle agitation). After rinsing with MilliQ-water the slides are ready for use. Long term storage for more than one year at 4°C (under air atmosphere) doesn't affect the substrates.
DNA microarray fabrication
Oligonucleotide microarrays tailor-made for our experiments were fabricated in-house employing light-directed in situ synthesis [14, 15]. The design of DMD based synthesis apparatus [16–21, 40] is described in Naiser et al. . Microarrays were synthesized in situ on hydroxy-functionalized phosphorus dendrimer supports. The initial photoreactive monolayer is created by coupling of NPPOC-dT-phosphoramidite. Subsequent light-directed synthesis was performed with NPPOC-phosphoramidite chemistry .
Probe sets for the experiments are derived from various 16–25 mer probe sequence motifs that are complementary to the set of fluorescently labeled target sequences (Tab. 1) available for this study. On the DNA chip each probe set (comprising between 64 and 400 features) is arranged as a closely spaced feature block (see Additional file 9) which during the analysis can easily be imaged as a whole. Compact arrangement reduces position-dependent systematic errors that can originate from gradients introduced during synthesis and/or hybridization (see below).
DNA chips produced for this study typically comprise about 2000 to 3000 features. A relatively large feature size of 21 μm (6 × 6 DMD pixels) is used to minimize image analysis related quantification errors.
Oligonucleotide target hybridization on the microarray – measurement of the hybridization signal intensity
Hybridization of fluorescently labeled targets to surface-bound probes is carried out in a temperature-controlled hybridization chamber. The chip, synthesized on a 20 mm diameter cover glass (glue-fixed onto a stainless steel support), constitutes a window into the chamber. The chamber volume of 150 μl is formed by a cutout in a 1.5 mm sheet of PDMS silicone rubber. Temperature is controlled with a foil heater attached to a stainless steel plate composing the backside of the hybridization chamber.
Relative intensities within the probe sets are largely independent of the hybridization time, chosen to be 10 minutes, typically. Probe sequence motifs with small hybridization affinities are hybridized for up to 30 minutes to achieve a sufficiently large hybridization signal/background ratio. Microarray hybridizations Hybridization temperature for 16 mer probes was typically 30°C. An increased hybridization temperature of 40°C has been applied for probes complementary to the target URA. At 30°C these, due to their large hybridization affinity, hybridize with reduced defect discrimination. Probes with a length of 20 and more bases are hybridized at 40°C. Hybridization is monitored in real-time on an Olympus IX81 fluorescence microscope. During acquisition of the hybridization signal the microarray is left in the hybridization solution. A 10 × 0.4NA UPlanApo objective provides a sufficiently large field of view. An electron multiplying CCD camera (Hamamatsu EM-CCD 9102) with a 1000 × 1000 pixel resolution is used for image acquisition. During image acquisition shade correction is performed to compensate for intensity inhomogeneities in fluorescence excitation.
Image analysis software developed in-house is employed to read the intensities of hundreds of features simultaneously.
Hybridization signal analysis – normalization
Hybridization signal measurements are performed with the microarray immersed in the hybridization solution. Thus, the measured hybridization intensity signal Ifeat,measis composed of the feature intensity I feat and the solution background intensity I back (originating from fluorescent targets floating above the microarray in the hybridization solution). The overall intensity Ifeat,meas= f (x)·(I feat + I back ) is affected by the function f(x) which accounts for spatial variations of the fluorescence excitation and the light collection efficiency of the microscope system (e.g. due to vignetting). Apart from Ifeat,measwe also locally (i.e. next to the corresponding microarray feature – see Additional file 10A) measure the solution background intensity f(x)·I back . A solution-background correction is performed by subtraction of the background fluorescence intensity. Further, by division by the solution background intensity f(x)·I back we cancel the feature-position related bias f(x).
In the further analysis we separate between the relatively strong defect positional influence and the defect-type related influence on the binding affinity. The positional influence is calculated as the moving average of mismatch hybridization signals (including all mismatch types) over a window of five consecutive MM-positions. By subtraction of the mean profile we obtain the MM-type dependent contributions δI MM to the hybridization signal.
To compare δI MM from different defect profiles it is necessary to account for the fact that the mismatch discrimination depends on the binding affinity. Mismatch discrimination is stronger in weakly-binding short duplexes or duplexes with a large AT-content. Vice versa, in case of duplexes with larger binding affinities the differences between PM and MM duplexes and among different MMs, respectively, may be rather small. We performed normalization of δI MM by division by the standard deviation σ profile (see Additional file 10B), or, alternatively, by division by the average of all MM hybridization signals of the corresponding MM defect profile.
Design of the DNA chip experiments
The flexibility of the in situ synthesis and the excellent spot homogeneity simplifies a comprehensive comparative analysis with the capability to detect subtle differences of the probe binding affinities. The experiments mainly differ in selection and spatial arrangement of the probe sequences. Particular experiments focus on the extraction of the positional dependence, the comparison of different defect types and on the identification of further influential parameters.
Spatial variations of the photodeprotection intensity and optical aberrations affecting the imaging contrast can result in gradients (as indicated in Additional file 11B) of the probe DNA quality (due to a varying number of synthesis errors). Thus, for a reliable determination of subtle differences in hybridization affinities, probes to be compared directly should be closely spaced on the microarray.
In the following we describe the design of the individual experiments:
Single base mismatch study
To investigate the positional dependence of single base mismatches and the impact of the mismatch type, we designed microarrays containing comprehensive sets of MM probes derived from a series of 25 16 mer probe sequence motifs. Position and type of the mismatch base pair were systematically varied, allowing us later to distinguish between the dominating positional dependence and other influential factors.
The features are arranged in groups of four (see Additional file 11A), corresponding to the four possible substituent bases (A, C, G and T) at a particular base position. A group comprises three mismatch probes plus one perfect match probe (PM) used for control. Sixteen of these feature groups (one for each base position) are arranged in a square feature block comprising in total 64 features (Additional files 9 and 11A).
Single base bulges
Single base insertions and deletions, due to an extra unpaired base result in bulged duplexes with reduced stability. A comprehensive study on the impact of single base insertions was performed using the chip design shown in Additional file 11A. The experiment comprised about 1000 single base insertion probes (insertion base type and position systematically varied) derived from twelve 20 to 25 mer probe sequence motifs.
Direct comparison of single base MMs and single base bulges
An experiment allowing for a direct comparison of PM, MM, single base insertion and deletion probes has been performed. Probe sets were derived from 16 mer probe sequence motifs, complementary to the targets listed in Tab. 1. For each of the 16 possible defect positions a set of 9 probes (comprising four single base insertions, one base deletion, three MMs and one PM probe) has been created. To avoid that a regular arrangement of the probe features could possibly affect the measurement (e.g. by introducing a bias due to increased target depletion near a PM probe), the sets of nine probes were randomly arranged in 3 × 3 matrices (Additional file 11B).
Direct comparison between DNA/DNA and DNA/RNA mismatches
The chip design (Additional file 11B) and the experimental procedures were basically identical with that of the previous experiment. Hybridization assays were conducted with fluorescently labeled DNA targets and corresponding RNA targets (Tab. 1). To avoid fabrication-related variation of the hybridization signals the hybridization assays were performed on the same chip, initially with RNA and subsequently, after regeneration of the microarray (by heating to 70°C in pure hybridization buffer), with the corresponding DNA targets.
Three microarrays were fabricated, each one focussing on one particular target sequence (COM, PET and LBE). Each microarray assay investigated single base MM and bulge defects for 6 different probe sequence motifs (obtained by shifting the 16 to 20 mer probe motif with respect to the longer target sequence). Two replicates of each feature block are employed to control for the reproducibility of the measurement.
Hybridization assays with the three microarrays were performed independently and on different days. The subsets of data obtained from the each of the assays display the same defect-type dependent trend for the defect-type dependent binding affinities. Yet smaller subsets from the individual defect profiles (originating from a single probe sequence motif) show basically the same trend of binding affinities which is, however, superposed by a strong sequence dependent bias.
This work was financially supported by the University of Bayreuth.
- Conner BJ, Reyes AA, Morin C, Itakura K, Teplitz RL, Wallace RB: Detection of sickle-cell beta-s-globin allele by hybridization with synthetic oligonucleotides. Proceedings Of The National Academy Of Sciences Of The United States Of America. 1983, 80: 278-282. 10.1073/pnas.80.1.278.View ArticleGoogle Scholar
- Schwarz DS, Ding HL, Kennington L, Moore JT, Schelter J, Burchard J, Linsley PS, Aronin N, Xu ZS, Zamore PD: Designing siRNA that distinguish between genes that differ by a single nucleotide. Plos Genetics. 2006, 2 (9): e140-10.1371/journal.pgen.0020140.View ArticleGoogle Scholar
- Wallace RB, Shaffer J, Murphy RF, Bonner J, Hirose T, Itakura K: Hybridization of synthetic oligodeoxyribonucleotides to phi-chi-174 DNA effect of single base pair mismatch. Nucleic Acids Research. 1979, 6 (11): 3543-3557. 10.1093/nar/6.11.3543.View ArticleGoogle Scholar
- Allawi HT, SantaLucia J: Thermodynamics and NMR of internal GT mismatches in DNA. Biochemistry. 1997, 36 (34): 10581-10594. 10.1021/bi962590c.View ArticleGoogle Scholar
- Peyret N, Seneviratne PA, Allawi HT, SantaLucia J: Nearest-neighbor thermodynamics and NMR of DNA sequences with internal AA, CC, GG, and TT mismatches. Biochemistry. 1999, 38 (12): 3468-3477. 10.1021/bi9825091.View ArticleGoogle Scholar
- SantaLucia J, Hicks D: The thermodynamics of DNA structural motifs. Annual Review of Biophysics and Biomolecular Structure. 2004, 33: 415-440. 10.1146/annurev.biophys.32.110601.141800.View ArticleGoogle Scholar
- Pozhitkov A, Noble PA, Domazet-Loso T, Nolte AW, Sonnenberg R, Staehler P, Beier M, Tautz D: Tests of rRNA hybridization to microarrays suggest that hybridization characteristics of oligonucleotide probes for species discrimination cannot be predicted. Nucleic Acids Research. 2006, 34 (9): e66-10.1093/nar/gkl133.View ArticleGoogle Scholar
- Urakawa H, El Fantroussi S, Smidt H, Smoot JC, Tribou EH, Kelly JJ, Noble PA, Stahl DA: Optimization of single-base-pair mismatch discrimination in oligonucleotide microarrays. Applied and environmental microbiology. 2003, 69 (5): 2848-2856. 10.1128/AEM.69.5.2848-2856.2003.View ArticleGoogle Scholar
- Wick LM, Rouillard JM, Whittam TS, Gulari E, Tiedje JM, Hashsham SA: On-chip non-equilibrium dissociation curves and dissociation rate constants as methods to assess specificity of oligonucleotide probes. Nucleic Acids Research. 2006, 34 (3): e26-10.1093/nar/gnj024.View ArticleGoogle Scholar
- Naiser T, Mai T, Michel W, Ott A: Versatile maskless microscope projection photolithography system and its application in light-directed fabrication of DNA microarrays. Review of Scientific Instruments. 2006, 77 (6): 063711-10.1063/1.2213152.View ArticleGoogle Scholar
- Luebke KJ, Balog RP, Garner HR: Prioritized selection of oligodeoxyribonucleotide probes for efficient hybridization to RNA transcripts. Nucleic Acids Research. 2003, 31 (2): 750-758. 10.1093/nar/gkg133.View ArticleGoogle Scholar
- Binder H: Thermodynamics of competitive surface adsorption on DNA microarrays. Journal of Physics-Condensed Matter. 2006, 18 (18): S491-S523. 10.1088/0953-8984/18/18/S02.View ArticleGoogle Scholar
- Bishop J, Blair S, Chagovetz AM: A competitive kinetic model of nucleic acid surface hybridization in the presence of point mutants. Biophysical Journal. 2006, 90 (3): 831-840. 10.1529/biophysj.105.072314.View ArticleGoogle Scholar
- Fodor SPA, Read JL, Pirrung MC, Land Lu Stryer, Solas D: Light-directed, spatially addressable parallel chemical synthesis. Science. 1991, 251 (4995): 767-773. 10.1126/science.1990438.View ArticleGoogle Scholar
- McGall GH, Barone AD, Diggelmann M, Fodor SA, Gentalen E, Ngo N: The efficiency of light-directed synthesis of DNA arrays on glass substrates. Journal of the American Chemical Society. 1997, 119 (22): 5081-5090. 10.1021/ja964427a.View ArticleGoogle Scholar
- Singh-Gasson S, Green RD, Yue YJ, Nelson C, Blattner F, Sussman MR, Cerrina F: Maskless fabrication of light-directed oligonucleotide microarrays using a digital micromirror array. Nature Biotechnology. 1999, 17 (10): 974-978. 10.1038/13664.View ArticleGoogle Scholar
- Gao XL, LeProust E, Zhang H, Srivannavit O, Gulari E, Yu PL, Nishiguchi C, Xiang Q, Zhou XC: A flexible light-directed DNA chip synthesis gated by deprotection using solution photogenerated acids. Nucleic Acids Research. 2001, 29 (22): 4744-4750. 10.1093/nar/29.22.4744.View ArticleGoogle Scholar
- Nuwaysir EF, Huang W, Albert TJ, Singh J, Nuwaysir K, Pitas A, Richmond T, Gorski T, Berg JP, Ballin J, McCormick M, Norton J, Pollock T, Sumwalt T, Butcher L, Porter D, Molla M, Hall C, Blattner F, Sussman MR, Wallace RL, Cerrina F, Green RD: Gene expression analysis using oligonucleotide arrays produced by maskless photolithography. Genome Research. 2002, 12 (11): 1749-1755. 10.1101/gr.362402.View ArticleGoogle Scholar
- Luebke KJ, Balog RP, Mittelman D, Garner HR: Digital optical chemistry: A novel system for the rapid fabrication of custom oligonucleotide arrays. Microfabricated Sensors, Application of Optical Technology for DNA Analysis. Edited by: Richard Kordal, Author Usmani, Wai Tak Law. 2002, American Chemical Society Publications, 815: 87-106.View ArticleGoogle Scholar
- Cerrina F, Blattner F, Huang W, Hue Y, Green R, Singh-Gasson S, Sussman M: Biological lithography: development of a maskless microarray synthesizer for DNA chips. Microelectronic Engineering. 2002, 61-2: 33-40. 10.1016/S0167-9317(02)00566-X.View ArticleGoogle Scholar
- Baum M, Bielau S, Rittner N, Schmid K, Eggelbusch K, Dahms M, Schlauersbach A, Tahedl H, Beier M, Guimil R, Scheffler M, Hermann C, Funk JM, Wixmerten A, Rebscher H, Honig M, Andreae C, Buchner D, Moschel E, Glathe A, Jager E, Thom M, Greil A, Bestvater F, Obermeier F, Burgmaier J, Thome K, Weichert S, Hein S, Binnewies T, Foitzik V, Muller M, Stahler CF, Stahler PF: Validation of a novel, fully integrated and flexible microarray benchtop facility for gene expression profiling. Nucleic Acids Research. 2003, 31 (23): e151-10.1093/nar/gng151.View ArticleGoogle Scholar
- Suzuki S, Ono N, Furusawa C, Kashiwagi A, Yomo T: Experimental optimization of probe length to increase the sequence specificity of high-density oligonucleotide microarrays. BMC Genomics. 2007, 8: 373-10.1186/1471-2164-8-373.View ArticleGoogle Scholar
- Fish DJ, Horne MT, Brewood GP, Goodarzi JP, Alemayehu S, Bhandiwad A, Searles RP, Benight AS: DNA multiplex hybridization on microarrays and thermodynamic stability in solution: a direct comparison. Nucleic Acids Research. 2007, 35 (21): 7197-7208. 10.1093/nar/gkm865.View ArticleGoogle Scholar
- Zhu J, Wartell RM: The effect of base sequence on the stability of RNA and DNA single base bulges. Biochemistry. 1999, 38 (48): 15986-15993. 10.1021/bi9916372.View ArticleGoogle Scholar
- Kierzek R, Burkard ME, Turner DH: Thermodynamics of single mismatches in RNA duplexes. Biochemistry. 1999, 38 (43): 14214-14223. 10.1021/bi991186l.View ArticleGoogle Scholar
- Dorris DR, Nguyen A, Gieser L, Lockner R, Lublinsky A, Patterson M, Touma E, Sendera TJ, Elghanian R, Mazumder A: Oligodeoxyribonucleotide probe accessibility on a three-dimensional DNA microarray surface and the effect of hybridization time on the accuracy of expression ratios. BMC Biotechnology. 2003, 3: 6-10.1186/1472-6750-3-6.View ArticleGoogle Scholar
- Naef F, Magnasco MO: Solving the riddle of the bright mismatches: Labeling and effective binding in oligonucleotide arrays. Physical Review E. 2003, 68: 011906-10.1103/PhysRevE.68.011906.View ArticleGoogle Scholar
- Zhang L, Miles MF, Aldape KD: A model of molecular interactions on short oligonucleotide microarrays. Nature Biotechnology. 2003, 21 (7): 818-821. 10.1038/nbt836.View ArticleGoogle Scholar
- Mei R, Hubbell E, Bekiranov S, Mittmann M, Christians FC, Shen MM, Lu G, Fang J, Liu WM, Ryder T, Kaplan P, Kulp D, Webster TA: Probe selection for high-density oligonucleotide arrays. Proceedings of the National Academy of Sciences of the United States of America. 2003, 100 (20): 11237-11242. 10.1073/pnas.1534744100.View ArticleGoogle Scholar
- Binder H, Kirsten T, Loeffler M, Stadler P: Sequence specific sensitivity of oligonucleotide probes. Proceedings of the German Bioinformatics Conference 2003. 2003, 2: 145-147.Google Scholar
- Peterson AW, Wolf LK, Georgiadis RM: Hybridization of mismatched or partially matched DNA at surfaces. Journal of the American Chemical Society. 2002, 124 (49): 14601-14607. 10.1021/ja0279996.View ArticleGoogle Scholar
- Sugimoto N, Nakano M, Nakano S: Thermodynamics-structure relationship of single mismatches in RNA/DNA duplexes. Biochemistry. 2000, 39 (37): 11270-11281. 10.1021/bi000819p.View ArticleGoogle Scholar
- Allawi HT, SantaLucia J: Nearest neighbor thermodynamic parameters for internal GA mismatches in DNA. Biochemistry. 1998, 37 (8): 2170-2179. 10.1021/bi9724873.View ArticleGoogle Scholar
- Ke SH, Wartell RM: Influence of neighboring base-pairs on the stability of single-base bulges and base-pairs in a DNA fragment. Biochemistry. 1995, 34 (14): 4593-4600. 10.1021/bi00014a012.View ArticleGoogle Scholar
- Znosko BM, Silvestri SB, Volkman H, Boswell B, Serra MJ: Thermodynamic parameters for an expanded nearest-neighbor model for the formation of RNA duplexes with single nucleotide bulges. Biochemistry. 2002, 41 (33): 10406-10417. 10.1021/bi025781q.View ArticleGoogle Scholar
- Gibbs JH, Dimarzio EA: Statistical mechanics of helix-coil transitions in biological macromolecules. Journal of Chemical Physics. 1959, 30: 271-282. 10.1063/1.1729886.View ArticleGoogle Scholar
- Kittel C: Phase transition of a molecular zipper. American Journal of Physics. 1969, 37 (9): 917-&. 10.1119/1.1975930.View ArticleGoogle Scholar
- Rodriguez-Lebron E, Paulson HL: Allele-specific RNA interference for neurological disease. Gene Therapy. 2006, 13 (6): 576-581. 10.1038/sj.gt.3302702.View ArticleGoogle Scholar
- Le Berre V, Trevisiol E, Dagkessamanskaia A, Sokol S, Caminade AM, Majoral JP, Meunier B, Francois J: Dendrimeric coating of glass slides for sensitive DNA microarrays analysis. Nucleic Acids Research. 2003, 31 (16): e88-10.1093/nar/gng088.View ArticleGoogle Scholar
- Kim C, Li M, Venkataramaia N, Richmond K, Kaysen J, Cerrina F: DNA microarrays: an imaging study. Journal of Vacuum Science & Technology B. 2003, 21: 2946-10.1116/1.1627802.View ArticleGoogle Scholar
- Hasan A, Stengele KP, Giegrich H, Cornwell P, Isham KR, Sachleben RA, Pfleiderer W, Foote RS: Photolabile protecting groups for nucleosides: Synthesis and photodeprotection rates. Tetrahedron. 1997, 53 (12): 4247-4264. 10.1016/S0040-4020(97)00154-3.View ArticleGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.