Introduction

Bioconjugation of nucleic acids with diverse biomolecules as well as a variety of solid ligands is of great importance in medicine and bionanotechnology1,2. As such, oligonucleotide (ON) bioconjugates hold promise for targeted drug delivery3,4, bioimaging5,6, biosensing7,8, as well as for the development of DNA nanomaterials9,10 and DNA-based nanodevices11,12. Due to such a broad application range of DNA-based systems, a site-specific immobilization of DNA molecules and subsequent hybridization are the two major events to be developed and optimized. In addition, nucleic acid bioconjugation plays a significant role in elucidating many of cellular processes, mainly those involving protein-DNA/RNA interactions13. Thus, a precise knowledge of the contacts made by the transiently binding proteins and nucleic acid molecules is essential to unravel the mechanisms of protein-nucleic acid interaction as well as to understand the architecture of such assemblies.

The conjugation of nucleic acids to a ligand, whether it is a biomolecule or a solid support, is brought by the reaction of particular functional groups between the two targets. Many methods of nucleic acid labelling are available, which are generally based on linking a specific tag to a sugar phosphate backbone at either 3′- or 5′-end of a nucleic acid molecule. The ease and efficiency of DNA end functionalization using automated solid-phase synthesis dictates the choice of modification14. Meanwhile, the attachment of a particular group to the nucleobase is a much more attractive approach in terms of sequence-specific labelling and programmability. Apart from a greater repertoire of modifications, the base-alteration offers a discontinuous space for functionalization with highly reactive but rather sophisticated and unique substitutions, such as aldehyde, acrylamide, vinylsulfonamide, chloroacetamide, trifluoroacetophenone or click chemistry-based functional groups, that are suitable for various bioconjugation procedures15,16,17,18,19,20. Considering base-modified nucleotides, terminal deoxynucleotidyl transferase (TdT)-mediated 3′-end tailing is a very promising approach for the enzymatic preparation of functional DNA species21,22,23,24,25. In fact, a wide spectrum of base-modified DNA oligomers bearing reactive alkyne-, azide-, oxanine-, amine- or imidazole- moieties can be prepared during TdT-assisted tailing, and further applied for the bioconjugation with proteins, coupling with fluorophores or immobilization26,27,28,29.

Photochemical cross-linking is a powerful strategy for studying biomolecule interactions as well as for anchoring a variety of ligands on different surfaces. A general photo-induced covalent cross-linking procedure can be achieved either by using an external photolinker30 or by introducing a native photoactive functional group into a targeted molecule31, the latter being a more specific and efficient approach. Aryl ketones, aryl azides and diazirines are the most popular photoreactive groups used for the photo cross-linking30,31. In fact, these have been used to study the cross-linking of DNA-DNA32 or protein-DNA/RNA complexes33,34,35,36, to probe protein-protein interactions37,38 or to attach a variety of biomolecules to polymer surfaces39,40. In addition, novel click chemistry-based photo-crosslinkers (e.g., tetrazole) have proved to be very promising as alternative photo-labels41.

Benzophenone (BP) is the most widely used and versatile photophore in bioorganic chemistry and material science42. In addition to an extraordinary stability compared to other cross-linking agents, BP is activated in a 350–365 nm UV range, which avoids protein and nucleic acid damage. BP and its derivatives are widely used for the photo-cross-linking of polymers as well as for the modification of both two- and three-dimensional solid supports42. In addition, BP-induced photochemical immobilization, grafting and patterning enable generation of biocompatible materials43,44,45 or 3D scaffolds46,47. Such functionalized platforms can be further utilized for the construction of diverse molecular diagnostic tools, such as DNA microarrays48,49. Furthermore, BP-based photoprobes are very frequently used to map interactions between the two molecules42. It has been demonstrated that the interaction between proteins can be detected using a BP-containing amino acid that can be either genetically encoded37 or incorporated via solid-phase synthesis50. Similarly, protein-DNA complexes can be identified using a BP-bearing peptide51. Nevertheless, the most common approach to a BP-mediated photo-cross-linking is the application of BP as a discrete photoinitiator52. Regarding BP-induced photoimmobilization strategies, the solid support itself must be coated, chemically activated or otherwise pre-treated, thus limiting its utilization. It seems that despite a huge variety of BP-containing cross-linking agents, the time-consuming synthetic routes render its applications inflexible.

Recently, we have described a set of N4-modified 2′-deoxycytidine 5′-triphosphates, including acetophenone (AP) and BP moieties-bearing nucleotide analogues53. We have found that AP- and BP-containing cytidine nucleotides are efficiently incorporated by a variety of family A and B DNA polymerases. In addition, we have shown that TdT is able to generate single-stranded DNA (ssDNA) fragments bearing up to several hundred such functional groups. Here, we report on the application of AP- and BP-functionalized DNA for the UVA-induced cross-linking with interacting proteins and describe an AP- and BP-mediated DNA photoimmobilization on various untreated solid supports. The results may prove useful for both the development of a feasible and direct DNA photoimmobilization approach as well as for the discovery of novel DNA-protein interactions.

Results

UVA-induced photo-cross-linking of AP- or BP-bearing ONs to TdT

Our strategy for the synthesis of reactive ON photoprobes bearing AP- or BP-functional groups exploited the TdT-based template-independent 3′-elongation. A set of N4-modified cytidine triphosphates was used, comprising a series of AP- or BP-containing nucleotide isomers (Fig. 1). TdT-mediated 3′-end tailing using the aforementioned cytidine analogues was described previously53. To examine a feasibility of AP- or BP-induced covalent cross-linking of modified DNA with a DNA-binding protein, we have chosen TdT as a particularly relevant model for the 3′-end-specific DNA-binding. This enabled us to simplify the process and perform a photo-cross-linking procedure immediately after the completion of elongation reactions.

Figure 1
figure 1

Chemical structures of N4-modified dCTPs used in this study. dCoAPTP: N4-(2-acetyl-benzoyl)-2′-deoxycytidine-5′-triphosphate, dCmAPTP: N4-(3-acetyl-benzoyl)-2′-deoxycytidine-5′-triphosphate, dCpAPTP: N4-(4-acetyl-benzoyl)-2′-deoxycytidine-5′-triphosphate; dCoBPTP: N4-(2-benzoyl-benzoyl)-2′-deoxycytidine-5′-triphosphate, dCmBPTP: N4-(3-benzoyl-benzoyl)-2′-deoxycytidine-5′-triphosphate, dCpBPTP: N4-(4-benzoyl-benzoyl)-2′-deoxycytidine-5′-triphosphate.

Full size image

Figure 2 shows the formation of covalent conjugates after exposure to a 365 nm UV light, where DNA-TdT cross-links are represented by the bands with lower mobility. Since a one-pot photochemical cross-linking approach was utilized, a mixture of DNA photoprobes of different length caused the formation of several conjugates, explaining the multiple bands present. In general, the yield of the cross-linking strongly depended on the number of photoreactive groups present on a DNA molecule. Since both para-isomers were revealed to be better substrates for TdT than ortho- or meta-analogues53, the cross-linking using pAP- or pBP-containing DNA was the most efficient (Fig. 2B). Moreover, the addition of an excessive amount of TdT led to an even more effective formation of covalent complexes. If the length of the modified DNA must be constant, or the amount of the targeted protein is limiting, one can prolong the time of irradiation rather than vary the critical parameters.

Figure 2
figure 2

UVA-induced cross-linking of AP- or BP-containing DNA to TdT. (A) A scheme of the photo-cross-linking between 3′-modified DNA and DNA-binding protein upon exposure to UV365 nm light. (B) Autoradiograms of SDS-PAGE showing covalent bioconjugates of the indicated 3′-modified ONs and TdT. The used nucleotides are indicated above the lanes.

Full size image

It was noticeable that after UV irradiation, a considerable part of material was unable to enter the gel, particularly in the case of pAB or pBB modifications (Fig. 2B). To determine its origin, the irradiated samples were further treated with protease. The disappearance of all the bands with the lower mobility (including the ones at the wells) demonstrated that all covalent adducts formed were DNA-protein rather than DNA-DNA conjugates (Fig. 2B). Due to a high number of photoreactive as well as highly hydrophobic groups present in a nucleic acid molecule, formation of multiple covalent bonds or specific nanostructured aggregates, or both, could occur.

To investigate the specificity of the cross-linking, several control reactions were done. Therefore, after the completion of elongation reactions, TdT was removed following the addition of other protein targets and subsequent exposure to UV radiation. To show that AP/BP modification-containing DNA can be cross-linked not only with TdT but with other DNA binding proteins as well, we attempted to couple modified ON with a single-stranded DNA binding protein (SSB) as a representative. Native-PAGE revealed formation of several dCpBP-ON:SSB cross-linked products that can be explained by the heterogeneity of the modified DNA as well as the multimericity of SSB (Supplementary Fig. S1). The requirement of a specific DNA-protein interaction for a successful cross-linking was also confirmed by the negative control experiment utilizing BSA where no covalent adducts were detected (Supplementary Fig. S1).

Photoimmobilization of N 4-modified DNA onto solid supports

Next, we performed a covalent attachment of ssDNA, via its modified 3′-mAP- or mBP-containing tail, to a variety of intact solid substrates. Polystyrene (PS), polypropylene (PP), polylactate (PLA), polydimethylsiloxane (PDMS), borosilicate (BS) glass and sol-gel based composites served as model polymer supports. Accordingly, the untreated ON as well as ssDNA oligomers 3′-extended using a natural dCTP were used as the negative control.

To investigate the dependence of the photoimmobilization efficiency on the number of photoreactive groups present in a nucleic acid molecule, the duration of the 3′-tailing reaction was varied. In fact, ranging the duration of TdT-mediated elongation from 10 sec to 60 min, several to several hundred of photoinducable nucleotides were appended, respectively (Fig. 3A). In the case of BP-modified DNA products, a considerable portion of such radioactive material was unable to enter the gel (Fig. 3A). Performing elongation reactions in the dark did not showed any changes (data not shown). Surprisingly, this bulky material was unable to enter the agarose gel also (Supplementary Fig. S2). In fact, to mobilize such large DNA, pulse-field gel electrophoresis was performed, yet we did not observe any migration of the modified DNA whatsoever (Supplementary Fig. S3).

Figure 3
figure 3

Photoimmobilization of AP- or BP-containing DNA to a PS support. (A) An autoradiogram of denaturing polyacrylamide gel showing TdT-catalysed 3′-end tailing with dCTP, dCmAPTP and dCmBPTP as substrates. The used nucleotides and elongation time are indicated above the lanes. (B) An autoradiogram of UVA-induced photoimmobilization of dCmAP-ONs and dCmBP-ONs to PS. The relative percentage of immobilization is indicated at the bottom where the most efficient immobilization was chosen to serve as 100%. Data represents the average of three independent experiments. (C) A schematic representation of AP- or BP-induced DNA photoimmobilization.

Full size image

The direct impact of the amount of AP- or BP-moieties on DNA attachment efficiency is illustrated in Fig. 3. Clearly, the more photoactive modifications (either AP or BP) that were present on an ON, the better such ON was attached to a PS substrate. The data presented in Fig. 3 indicate that a BP-functionalization induced a more effective immobilization. Indeed, only a few BP-modifications gave superior immobilization results compared to those observed for a large number of AP-groups. Neither untreated ON, nor 3′-dC-extended DNA fragments showed any indications of immobilization (Fig. 3).

Further, a successful photoimmobilization of AP- or BP-modified DNA was accomplished using PP, PLA, PDMS, BS as substrates as well as metal-containing (Al, Ti, Zr or V) sol-gel based hybrid materials (Fig. 4). In general, both AP- and BP-containing DNA demonstrated similar UVA-induced immobilization efficiencies on every solid substrate tested.

Figure 4
figure 4

Autoradiograms of UVA-induced photoimmobilization of dCmAP-ONs and dCmBP-ONs to different solid supports. The used photoreactive probes and solid supports are indicated above and below each autoradiogram, respectively.

Full size image

Hybridization of DNA targets to photoimmobilized BP-containing DNA probes

To verify that the attached 3′-dCBP-tailed DNA probes are properly oriented for a successful detection of complementary DNA, we performed a target–probe hybridization procedure. To detect only specific DNA sequences, two different ssDNA probes P1 and P2 were chosen. Both were elongated at the 3′-end using TdT and dCpBPTP, and photoimmobilized onto PS plates. A 5′-radiolabelled ssDNA fragment P1R served as a DNA target. P1R sequence perfectly matched the DNA probe P1, whilst the thermal stability of P1R-P2 dimer was extremely low. Therefore, a hybridization signal upon formation of P1-P1R dimer represented a positive detection system.

To determine targets with different sequence similarity, various hybridization conditions were tested. High stringency hybridization conditions (e.g., high temperature, low salt concentration) permit hybridization only between highly similar nucleic acid molecules, whereas low stringency conditions (low temperature, high salt) allow unspecific hybridization. Remarkably, our results demonstrated that even under low stringency hybridization conditions (20 °C, ~0.8 M salt), hybridization occurred specifically between complementary sequences (Fig. 5B). Moreover, the detection of complementary DNA targets was possible in a wide temperature range (20–55 °C) as well as within short time periods (2 hours).

Figure 5
figure 5

Hybridization of DNA targets to photoimmobilized 3′-dCpBP-DNA probes on a PS support. (A) A scheme visualizing hybridization experiment. (B) An autoradiogram of hybridization of 5′-radiolabelled ssDNA targets to photoimmobilized BP-bearing complementary DNA probes. (C) An autoradiogram of DNA denaturation of 5′-radiolabelled DNA targets. The used hybridization and denaturation conditions are indicated.

Full size image

Finally, to examine the suitability of 3′-dCBP-tailed DNA probes for the construction of DNA microarrays, we performed a facile procedure for DNA denaturation. To avoid harsh denaturation methods, such as alkaline or dimethyl sulfoxide treatment, we tested the most common and secure way to denature double-stranded-DNA (dsDNA) (particularly for PCR) – heating at high temperature (e.g., 95 °C). The samples were heated for different periods of time resulting in a mild denaturation after a brief heating, and full DNA melting after prolonged incubation (Fig. 5C). Overall, our results indicated that BP-containing DNA may become a promising alternative for a delicate and low-cost photoimmobilization, hybridization as well as denaturation.

Discussion

A strict temporal and spatial control provided by the UV-triggered formation of covalent junctions renders photo-induced cross-linking a powerful tool for the examination of biomolecular interactions. In addition, it is important that the UV cross-linking can be used not only for the in vitro studies but also for the capture of complexes in vivo, without perturbing their interface. One way to minimize a cross-linking heterogeneity, is a site-specific incorporation of the photoreactive functionality that readily localizes unique cross-linking contacts. Therefore, a considerable emphasis must be placed on the development of a simple as well as flexible approach for the functionalization of interacting partners. Aryl ketones (i.e., AP and BP) are very effective UV-inducible functional groups that upon irradiation generate a reactive triplet biradical, which predominantly abstracts a hydrogen atom from a donor molecule42. Aliphatic hydrogens or hydrogens adjacent to hydroxyl groups can serve as H-donors, depending on H steric accessibility and close proximity to AP/BP. As such, both proteins and nucleic acids, as well as a wide variety of organic and inorganic polymeric materials, may be subjected to a successful AP/BP-mediated cross-linking. Herein, we present a facile and straightforward approach to covalently couple the AP/BP-functionalized nucleic acid molecules with both proteins and diverse solid supports. We show that ssDNA oligomers 3′-tailed with AP- or BP-containing N4-modified cytidine nucleotides act as a photoprobe which is easily connected to DNA-binding proteins only upon exposure to UV light. We also observe that the efficiency of the cross-linking strongly depends on the quantity of photoactive groups present in a nucleic acid molecule. In fact, it makes ortho-AP/BP the least, whilst para-AP/BP isomers the most effective cross-linking agents. Our results highlight three important findings. First, a TdT-catalysed incorporation of dCAP and dCBP suggests that it is adjustable to any nucleic acid molecule regardless of its sequence context or length. Despite the presence of several alternatives for the enzymatic 3′-end labelling using other terminal transferase activity-bearing human polymerases (e.g., θ, λ or µ)54,55,56 or a polymerase acting as a ribozyme57, TdT is the only researched and marketed enzyme so far. Considering that TdT is known to extend ss- or ds- DNA58,59, RNA substrates60 as well as to perform de novo synthesis of polynucleotides61, the proposed method is of paramount importance since it can be easily adopted for the cross-linking of a wide spectrum of nucleic acids. Second, it is convenient because the number of photoreactive groups to be introduced can be easily varied. Third, irradiation at 365 nm wavelength, which can be referred to as UVA light (315–400 nm), causes absorbance by photoactive AP/BP moieties mostly, preventing direct damage and leaving native biomolecules intact.

A coupling of nanomaterials with biomolecules for the generation of functional nanosystems or nanodevices is yet another field where photo-cross-linking can emerge as a very promising approach. DNA immobilization and hybridization play a major role in the era of nanostructures, therefore all challenges and drawbacks of various immobilization procedures need to be acknowledged. Surface modification, functionalization or coating may be considered as necessary steps for the successful DNA immobilization1,62. Consequently, complicated surface preparation techniques affect the experimental design and limit the choice of relatively inert polymeric supports. PS, PP and BS are among the most popular biologically inert polymers, and are used in many applications. Glass substrates, in particular, are very often utilized for DNA microchips, however surface pre-treatment is necessary to attach biomolecules63. Our results offer an effortless way to directly immobilize any type of DNA onto a wide range of polymer substrates, including inert PS, PP and BS, by applying UV irradiation. In addition, we show that biodegradable polymers, such as PLA and PDMS, as well as the sol-gel based hybrid materials containing aluminium, vanadium, titanium or zirconium ions can be readily utilized as bioactive supports for DNA immobilization. Both PLA and silicone-based organic elastomer PDMS are widely used in targeted drug delivery and release systems, for detection of pathogens, as antisense therapeutics or as scaffolds in tissue engineering64. Similarly, sol-gel based composites show promise as nanocarriers, sensing devices or scaffolds for tissue repair65. Owing to the low cost, simple design and biocompatibility of such polymers, the functionalization will allow to expand their utilization in future.

Our findings indicate that the greater number of modified residues is present, the more efficient immobilization of such DNA molecules takes place. Remarkably, the largest 3′-tailed DNA products exhibit particularly abnormal electrophoretic mobility or no detectable mobility at all. Performing TdT-elongation in the dark did not prevent from accumulation of the material in or near the wells. Moreover, these aggregates were not digested by protease, which implied that an unknown material was not a DNA-TdT complex formed due to the inevitable exposure to UV present in a daylight. Altogether, it might be speculated that such phenomenon can be caused by the presence of a highly hydrophobic 3′-AP/BP-tail that seems likely to determine the formation of specific steric structures, aggregates or even nanoscaled particles that prevent from fractionation by conventional methods. Further, we suggest that the formation of covalent bonds between 3′-modified DNA and the solid support occurs at multiple sites rather than at a single 3′-end. Hence, a greater rigidity and stability of immobilized biomolecules is obtained which, in fact, directs DNA properly for the successful hybridization. Thermal denaturation, the most frequent and simple method of DNA denaturation, further ensures compatibility of the immobilized N4-acyl-dC-DNA probes for DNA hybridization based assays. Taken together our results show that a novel immobilization method presented here (i) does not require any prior chemical derivatization of solid supports, (ii) is suitable for a broad range of nucleic acid molecules (ss- or ds-DNA/RNA) as well as the most classic hybridization-denaturation procedures, and (iii) may prove useful for the rapid detachment of DNA from the polymeric support due to a certain pH-lability of N4-acyl-modifications.

In conclusion, a simple but effective photo-cross-linking approach based on the N4-modified-DNA bearing photoactive AP/BP modifications was designed and studied. 3′-tailed functionalized nucleic acid molecules containing several to several hundred AP or BP moieties were cross-linked with both an interacting protein and a variety of solid supports upon exposure to UV light. Successful DNA immobilization, hybridization as well as denaturation procedures were also achieved. Our findings indicate that N4-modified cytidine nucleotides bearing functional AP or BP groups can be considered as suitable building blocks for the construction of functionalized DNA for the UV-induced cross-linking.

Methods

Materials

All chemicals were of the highest reagent grade, unless otherwise specified. Synthetic ONs were purchased from Metabion (Planegg/Steinkirchen, Germany); sequences of the ONs are shown in Table 1. [γ-33P]-ATP was obtained from Perkin Elmer (Singapore). TdT, T4 polynucleotide kinase (T4 PNK), solutions of 2′-deoxyribonucleoside triphosphates, Zeba Spin desalting columns and BS glass slides were purchased from Thermo Fisher Scientific Baltics (Vilnius, Lithuania). The UV light source was purchased from Epileds (Tainan, Taiwan).

Table 1 Sequences of the ONs used in this study.
Full size table

Photo-cross-linking of 3′-dCAP- or 3′-dCBP-tailed ONs to TdT

The cross-linking apparatus was constructed as described previously, with slight modifications66. The apparatus consisted of an ice container, a 96-well plate, a sheet of parafilm and an UV light source. A sheet of parafilm was placed over the top of a 96-well plate, and taped to the plate on all four sides. Each well was pressed to create a shallow groove. The plate was kept on ice before and during irradiation. Samples were irradiated at 365 ± 5 nm (200–220 mW/cm2) 5 mm away from the surface of the light source, which provided dose of UV irradiation of ~8 J/cm2.

UV cross-linking of dCAP- or dCBP-elongated primer P1 to TdT was carried out after TdT-catalysed 3′-elongation reactions. The primer P1 was 5′-33P-labelled by using T4 PNK in the presence of [γ-33P]-ATP. The 5′-labelled primer was desalted using Zeba Spin desalting columns (7 K MWCO). The reaction mixtures (40 µL) consisted of TdT (0.5 U), 5′-33P-labelled P1 (5 nM), dCTP, dCAPTP or dCBPTP (10 µM) and glutamate reaction buffer (20 mM sodium glutamate, 20 mM NaCl, 10 mM DTT, 0.5% Triton X-100, 1 mM MgCl2 (pH 8.2)). The reaction mixtures were incubated for 5 min at 37 °C. Immediately after incubation, the reaction mixtures were chilled on ice and transferred as 10 µL drops to the wells on the parafilm tape. The ice container was placed underneath a 365 nm UV, and samples were irradiated for 5 min. Then the samples were transferred from parafilm wells to microtubes and quenched with 2 × loading solution. To verify the TdT-ON cross-links generated by the irradiation, the samples were supplemented with SDS loading dye, heated at 95 °C for 5 min, and analysed by electrophoresis on a 14% (w/v) SDS-PAGE gel. The cross-linking products were then visualized by phosphorimaging.

Preparation of solid supports for photochemical immobilization

PS and PP supports were cut out of a standard sterile Petri dish and a microcentrifuge tube, respectively. PLA polymer support was received as a gift from Egidijus Šimoliūnas and was manufactured as described previously67. PDMS and metal-containing sol-gel composite (hybrid) supports were received as a gift from Evaldas Balčiūnas. PDMS films were prepared by mixing the silicon base (Sylgard 184, Dow Corning) with a proprietary curing agent at 10:1 mass ratio in a glass Petri dish, followed by degassing under vacuum, curing for 1 hour at 100 °C, and finally removing the film. Hybrid material films containing aluminium, titanium, zirconium or vanadium were prepared on BS slides as described previously68,69,70. All polymer supports were prepared as 20 × 7 mm slides, washed with 200 µL of ethanol and air-dried.

Photochemical immobilization of 3′-dCAP- or 3′-dCBP-tailed ONs onto the solid supports

To immobilize dCAP- or dCBP-elongated ssDNA onto the solid supports, TdT-catalysed 3′-tailing was performed as described above, except for using a TdT buffer (supplied by the manufacturer of the enzyme). 5′-33P-radiolabelled ON P1 was used for the elongation and immobilization. Briefly, the reaction mixtures (150 µL) were incubated from 10 sec to 60 min at 37 °C, 10 µL samples were taken from the reaction mixture at predetermined times (10, 15, 30, 45 sec and 1, 5 and 60 min) and terminated with 1 µL of EDTA (200 mM) following a complete inactivation of TdT by heating (10 min 70 °C). After TdT inactivation, the reaction mixtures were chilled on ice and transferred as 2 µL drops onto a slide of solid support. Then the specimens were placed under the UV light source (5 mm away from the surface of the light source) and irradiated for 5 min (365 ± 5 nm). This provided dose of UV irradiation of ~8 J/cm2. Immediately after irradiation, the specimens were rinsed with 500 µL of wash buffer (50 mM potassium phosphate, 1% Triton X-100 (pH 7.0)) following incubation (1 h at room temperature) in 1.5 mL of wash buffer vigorously shaking. Then the specimens were rinsed with 200 µL of distilled water and air-dried. Photochemical immobilization of modified ONs to solid supports was then visualized by phosphorimaging. To determine the dependency of the UVA-induced immobilization on the length of the 3′-modified tail, the intensities of the spots representing the immobilized products were determined using OptiQuant analysis software (version 03.00, Packard Instrument Company Inc., Meriden, CT, USA).

Hybridization of ssDNA targets to the immobilized 3′-dCBP-tailed complementary DNA probes

A linear ssDNA probe P1 or P2 was 3′-tailed with dCpBP using TdT. The reaction mixture was prepared in a total volume of 10 µL and consisted of TdT (10 U), P1/P2 (500 nM), dCpBPTP (100 µM) and TdT buffer supplied by the manufacturer. The reaction mixtures were incubated for 10 min at 37 °C, quenched by the addition of 1 µL of EDTA (200 mM) following a complete inactivation of TdT by heating (10 min 70 °C). After TdT inactivation, the reaction mixtures were chilled on ice and transferred as 2.5 µL drops onto a slide of PS support. Then the specimens were placed under the UV light source (5 mm away from the surface of the light source) and irradiated for 5 min (365 ± 5 nm), which provided dose of UV irradiation of ~8 J/cm2. Immediately after irradiation, the specimens were rinsed with 500 µL of wash buffer (50 mM potassium phosphate, 1% Triton X-100 (pH 7.0)) following incubation (1 h at room temperature) in 1.5 mL of wash buffer vigorously shaking. Then the specimens were rinsed with 200 µL of distilled water and air-dried. Silicone was used to create a ring-fence around the area of immobilized ONs, forming a shallow groove. A primer P1R was 5′-radiolabelled as described above and diluted in a 5 × saline-sodium citrate hybridization buffer (0.75 M NaCl, 75 mM sodium-citrate, 0.02% SDS (pH 7.0)) to a final 10 nM concentration. The 5′-labelled DNA target P1R was transferred as 10 μL drops onto the immobilized DNA probes P1/P2, enclosed with silicone, and further sealed to prevent the evaporation. The specimens were incubated at room temperature, 37 °C or 55 °C, for 2 or 16 hours. The solution was removed from the wells, which were then washed six times with 1 mL of distilled water and air-dried. The DNA denaturation procedure was performed by repeatedly heating samples at 95 °C for 1, 5 or 15 min. Hybridization and denaturation results were then visualized by phosphorimaging.