DNA Strand Displacement with Base Pair Stabilizers: Purine‐2,6‐Diamine and 8‐Aza‐7‐Bromo‐7‐Deazapurine‐2,6‐Diamine Oligonucleotides Invade Canonical DNA and New Fluorescent Pyrene Click Sensors Monitor the Reaction

Abstract Purine‐2,6‐diamine and 8‐aza‐7‐deaza‐7‐bromopurine‐2,6‐diamine 2’‐deoxyribonucleosides (1 and 2) were implemented in isothermal DNA strand displacement reactions. Nucleoside 1 is a weak stabilizer of dA‐dT base pairs, nucleoside 2 evokes strong stabilization. Strand displacement reactions used single‐stranded invaders with single and multiple incorporations of stabilizers. Displacement is driven by negative enthalpy changes between target and displaced duplex. Toeholds are not required. Two new environmental sensitive fluorescent pyrene sensors were developed to monitor the progress of displacement reactions. Pyrene was connected to the nucleobase in the invader or to a dendritic linker in the output strand. Both new sensors were constructed by click chemistry; phosphoramidites and oligonucleotides were prepared. Sensors show monomer or excimer emission. Fluorescence intensity changes when the displacement reaction progresses. Our work demonstrates that strand displacement with base pair stabilizers is applicable to DNA, RNA and to related biopolymers with applications in chemical biology, nanotechnology and medicinal diagnostics.


Introduction
DNA strand displacement is a process in which one strand of duplex DNA is replaced by an invader strand, thereby releasing an output strand from the DNA double helix. The reaction takes place at constant temperature, is thermodynamically driven, and controlled by a negative ΔG value. Displacement occurs in many biological contexts and technological applications; in gene regulation, dynamic DNA nanotechnology, biosensing, and has led to the construction of DNA motors. [1] The heart of strand displacement is hybridization and the stronger binding of the invader strand to the target duplex compared to the released strand. Additional base pairs, so-called toeholds were implemented to stabilize the invader complex. [2] Another option to strengthen the binding of the invader strand to duplex DNA is the use of modified nucleosides in place of canonical nucleosides evoking higher base pair stability. [3] This process does not require toeholds. Toehold-free displacement was recently reported for anomeric DNA. [3d] In this protocol, different stabilities of heterochiral vs. homochiral DNA were applied.
In the realm of DNA, a number of modified nucleosides have been reported that stabilize Watson-Crick base pairs. [4] Some of them are naturally occurring; most are of synthetic origin. Nucleobases, sugar moieties, exocyclic substituents and the phosphate residue are common targets for functionalization. Chemistry, biology, medicinal diagnostics and therapeutics and more recently nanotechnology make use of this phenomenon. [5] Our laboratory has reported a series of compounds that are in common use for DNA detection, sensing, sequencing and the construction of entirely new DNAs. [4b,c,6] The purine-2,6-diamine nucleoside 1 [7] is a constituent of the Cyanophage S-2 DNA and replaces dA in the virus genome to 100 % (Figure 1). [7a] This nucleoside is a weak DNA stabilizer when it substitutes dA in the dA-dT base pair (purine numbering is used throughout the results and discussion section and systematic numbering in the Experimental Section, Figure 1). [8] With respect to dA, compound 1 contains an additional amino group at the 2-position of the nucleobase. Related nucleosides such as 2-amino-7-deazaadenine and 2amino-8-aza-7-deazaadenine 2'-deoxyribonucleosides show similar properties. [4b,9] Contrary to 1, the 8-aza-7-bromo-7deazapurine-2,6-diamine 2'-deoxyribonucleoside 2 is a strong DNA stabilizer. [4b,10] Here, the purine ring is replaced by a pyrazolo [3,4-d]pyrimidine skeleton and the heterocycle carries the amino group at the 2-position as in 1. The presence of an additional bromo atom at the 7-position enhances base pair stability ( Figure 1). According to this, nucleoside 2 is more effective in duplex stabilization than nucleoside 1. Furthermore, compound 2 possesses a more stable glycosylic bond compared to the labile purine nucleoside 1. [10a] In this work, we use the stabilizing forces of 1 and 2 to drive DNA strand displacement. Toeholds are not required.
For strand displacements reactions, sensors have to be implemented that monitor the progress of the displacement reaction and provide information on their kinetics. For this, FRET quencher systems are commonly applied. [3d] Recently, our laboratory used the fluorescence of ethidium bromide fluorescence as external sensor. For this work, the two new environmental sensitive fluorescence pyrene sensors 4 and 5 were developed and were implemented in the displacement systems.

Design of strand displacement systems A and B and function of base pair stabilizers and sensors
Stabilization of DNA is beneficial for many applications in DNA research. For the purpose, in the toehold-assisted displacement systems additional base pairs are implemented strengthening the binding capacity of the invader strand. [2,3] In the systems described in this work, existing base pairs are stabilized by base pair stabilizers. Stabilizer 1 is a naturally occurring nucleoside and has a long history in DNA research. [7,11] Stabilizer 2 is of synthetic origin. To provide the option for labelling, the clickable linker nucleoside 3 was designed. Nucleoside 2 found widespread application as efficient DNA stabilizer without disturbing the DNA double helical structure. [8h,10] The stabilizing properties of 1 and 2 were used to strengthen homochiral and heterochiral DNA and to harmonize the stability of the dA-dT and dG-dC base pair. [8h,10] The stabilizers 1 and 2 have been already incorporated in DNA by solid-phase synthesis employing phosphoramidite chemistry. [8h,10,12] The phosphoramidite building block of 3 with a clickable linker is described in this work.
To monitor the progress of strand displacement reactions, pyrene fluorescence was utilized. To this end, the new pyrene sensors 4 and 5 were developed. In sensor 4 pyrene is linked to the nucleobase, in sensor 5 two proximal pyrene residues are connected to an abasic linker unit. The phosphoramidite of the bis-pyrene sensor 5 is described in this work, pyrene sensor 4 is accessible by click chemistry on oligonucleotides. Scheme 1 shows the principle of the strand displacement reaction for systems A and B using base pair stabilizers. In system A, a single pyrene residue is part of the invader, whereas a bis-pyrene linker is part of the target duplex in system B and is later part of the output strand. The dendritic pyrene linker forms an extension of the backbone at the 5'-site. The single pyrene in system A shows monomer emission, whereas excimer emission is observed for the bis-pyrene linker in system B. Stabilizers 1 and 2 are expected to act as the driving force for the displacement in both systems.

Syntheses of stabilizers and pyrene sensors for incorporation in DNA
Key compounds for the synthesis of modified invader oligonucleotides are the phosphoramidites 8-10, that are used together with standard phosphoramidites to synthesize 12-mer oligonucleotides used in the displacement studies. Phosphoramidites 9 and 10 have been already described by our laboratory; [8h,10a] the synthesis of phosphoramidite 8 is reported herein (Scheme 2).
[10b] Sonogashira cross-coupling performed with 6 in anhydrous DMF and excess of 1,7octadiyne afforded the side chain derivative 3 in 73 % yield. [13] Then, both amino groups of nucleoside 3 were protected. Protecting groups were carefully chosen, as the 2-and 6-amino groups differ in reactivity. For the related 7-bromo-2,6-diamino nucleoside 2 the combination of a formyl group for 2-amino protection and a dibutylamidine residue for the 6-amino group was found to be the optimal combination. [10a] The same strategy was used for compound 3. Reaction of 3 with N,N-dibutylformamide dimethylacetal gave the 2,6-bis-amidine nucleoside together with traces of the 2-formyl-6-amidine compound as detected by TLC (CH 2 Cl 2 /MeOH, 9 : 1). After evaporation, the reaction mixture was directly used without purification in the next step. Tritylation with DMTÀ Cl was performed under standard conditions. As reported for other nucleosides, [9b,10a] the amidine group at the 2-position is selectively hydrolyzed during the work-up procedure and the formyl derivative 7 was isolated in 51 % yield over 2 steps. Finally, phosphitylation of 7 with chloro(2-cyanoethoxy)(diisopropylamino)phosphine gave the phosphoramidite 8 in 77 %.
The phosphoramidite 15 containing two pyrene residues attached to a dendritic linker was prepared by the route shown in Scheme 3. Pyrene alkyne 13 [14] was clicked to the dendritic bis-azide 12 [15] in t-BuOH/H 2 O/THF with CuSO 4 /ascorbic acid as catalysts. The bis-pyrene click adduct 5 was purified by column chromatography and obtained in 51 % yield over 2 steps. Then, 4,4-dimethoxytritylation was performed (!14, 52 %). Finally, phosphitylation furnished phosphoramidite 15 (70 %). We became inspired to construct a bis-pyrene sensor by our own studies on 1,3-pronanediol linkers and on pyrene double click functionalization performed on tripropargylamine nucleosides and oligonucleotides. [16] Further, motivation came from the work of the Yamana group and Saito and earlier work of Letsinger. [17] Yamana developed a linker similar to ours that differ in the connectivity of the pyrene residues to the propanediole unit. [17b] As the synthesis of our bis-pyrene sensor is significantly shorter than that of Yamana, we took advantage of this matter.
All newly synthesized compounds were characterized by ESI-TOF mass spectra, extinction coefficients and by 1 H-, 13 C NMR spectra. 1 H-13 C correlated (HMBC and HSQC) NMR spectra were used to assign the 13 C NMR signals. For details, see the Experimental section, for data Tables S1, S2, Supporting Information and for spectra, Figures S5-S31, Supporting Information.

Synthesis of functionalized oligonucleotides and their duplex stability
For the construction of the strand displacement systems, a series of oligonucleotides were synthesized. Synthesis was performed on solid-phase with modified phosphoramidites 8-10 and 15 (bis-pyrene) and employed together with those of canonical DNA. Single and multiple incorporations of modified nucleosides were performed. Coupling yields were always higher than 95 %. After synthesis, oligonucleotides were cleaved from the solid support and deprotected in conc. 28 % aq. NH 3 at 55°C for 12 h. Detritylation was performed with 2.5 % dichloroacetic acid in CH 2 Cl 2 . Oligonucleotides were purified before and after detritylation by reversed-phase HPLC on a RP-18 column. HPLC purity profiles are documented in the Supporting Information ( Figure S3, Supporting Information). New oligonucleotides are shown in Table 1 together with their molecular masses determined by MALDI-TOF mass spectrometry.
In order to prove the applicability of modified oligonucleotides containing stabilizer 1 or 2 for the toehold-free displacement systems A and B, thermodynamic stability of duplexes was investigated. To this end, single-stranded oligonucleotides were hybridized, thermal melting curves were measured, T m values were determined and thermodynamic data were calculated (Tables 2, 3, Tables S3 and Figures S3,S4,Supporting Information). In all experiments, the single strand concentration was 5 μM and the buffer solution contained 100 mM NaCl, 10 mM MgCl 2 , and 10 mM Na-cacodylate (pH 7.0).
The unmodified duplex ODN-1 * ODN-2 that is commonly used in our laboratory was the basis for all further experiments in this work. It shows a T m value of 47°C (Table 2, left column). The T m value increases to 52°C when the output strand ODN-2 was displaced by the invader strand ODN-4 containing three 2,6-diaminopurine nucleosides 1 in place of three dA residues. When the same was performed with invader strand ODN-6 (stabilizer 2), the stability increase was significantly stronger (T m = 60°C, ODN-1 * ODN-6). Replacing the central stabilizer 2 by the octadiynyl nucleoside 3 (ODN-1 * ODN-7) gave a T m of 58°C. Functionalization of 3 with pyrene (!4) yielded additional stabilization (T m = 66°C, ODN-1 * ODN-8) (for details, see the Experimental section). Duplex stability depends on the number of incorporations and their position. Earlier, it was reported that stabilizer 2 shows favourable proton donor properties compared to the 2,6-diaminopurine nucleoside 1. Most likely the enhanced amide character of the 2-amino group of 2 compared to 1 results from the changes of electronic properties of the two ring systems (8-aza-7-deazapurine vs. purine) and the electronic contribution of the halogen atom. Changes in base stacking and hydration have to be also considered. The same can be considered for stabilizer 3 and the pyrene sensor 4. From T m values and thermodynamic data, it is obvious that multiple incorporations of the 2,6-diamino nucleosides 2, 3 and 4 fulfil the expectation as stabilizers of the dA-dT base pairs in the displacement system. According to the higher stability of modified duplexes containing nucleoside 2, a clear advantage is observed for nucleoside 2 with respect to 1. Apparently, the third hydrogen bond is stronger in the 2-dT base pair as in the 1-dT pair and develops the same stability as in the dG-dC pair. [8,10] The octadiynyl modified nucleoside 3 with a clickable side chain was used to introduce pyrene. The terminal bis-pyrene linker in duplex ODN-1 * ODN-13 induces only marginal but measurable stabilization. The T m values in the presence or absence of invaders or output molecules are very similar to those of the "duplexes only" (Table 3). Table 1. Synthesized oligonucleotides and their molecular masses determined by MALDI-TOF mass spectrometry.

Chemistry-A European Journal
Research Article doi.org/10.1002/chem.202202412 Table 2. T m values of oligonucleotide duplexes before and after addition of the invader strand containing 2,6-diamino nucleosides 1-4 or the bis-pyrene linker 5. [ [°C] Measured at 260 nm at a concentration of 5 μM + 5 μM single strand at a heating rate of 1.0°C/min in 100 mM NaCl, 10 mM MgCl 2 , and 10 mM Na-cacodylate (pH 7.0). [b] T m values were calculated from the heating curves using the program Meltwin 3.0. [18] [c] T m values were calculated from the heating curves after adding the corresponding invader strands with 5 μM concentration. The standard deviation for the T m values is � 0.5°C. For the formulas of compounds 1-5, see the legend to Table 3.

Isothermal displacement reaction with oligonucleotide invaders containing base pair stabilizers 1 and 2 and pyrene or bis-pyrene sensors
DNA strand displacement is a dynamic process in which a single-stranded invader DNA replaces a strand with the same or similar strand recognition in the duplex target. The displacement reaction starts usually at the termini of duplexes as terminal base pairs are less stable than internal pairs. [1][2][3] The reaction is driven by the enthalpy change between the substrate duplex and the displaced duplex. The kinetics depends on the size of the duplex, the invader itself, and on many other parameters such as strand concentration, temperature or ions present in solution. When the displacement is finished, the back reaction is unfavorable, when the free energy difference between substrate and displaced duplex is substantial.
When sensors are part of a displacement system, they affect the stability of the parent duplex. [19] Consequently, all thermodynamic and kinetic data obtained for common displacement systems including those of the often used FRET system do not display the real situation of the systems in the absence of sensors. This is also visible from stability data of Tables 2, 3 (duplex stabilities in the presence or the absence of pyrene sensors). Nevertheless, sensors are required to monitor displacement and fluorescence is the method of choice.
Herein, fluorescence changes of pyrene are used. [20] To this end, the fluorescence of single strands and duplexes decorated with pyrene residues were measured and compared. To this end, 5 μM (sensor 4) or 2.5 μM (sensor 5) solutions of single strands in buffer were prepared and fluorescence was measured before and after addition of complementary strands (Figure 2, Figure S2, Supporting Information). Invader oligonucleotides with sensor 4 show monomer emission around 384 and 398 nm, whereas oligonucleotides with the dendritic bis-pyrene linker 5 show excimer emission at 488 nm. The oligonucleotides were excited at 345 nm (sensor 4) and 347 nm (sensor 5).
From Figures 2a and 2b significant fluorescence changes between single strands and duplexes are apparent. This is the case for oligonucleotides with sensor 4 having the pyrene residue connected to the nucleobase as well as for oligonucleotides with the terminal bis-pyrene linker 5. According to this, their application in strand displacement was verified.
Next, the progress of the displacement reactions was followed by time-dependent recording of pyrene fluorescence change. On molecular level, the invader strand targets the fraying ends of the substrate duplex and displacement progresses by strand migration (Scheme 4). To this end, the invader single strand ODN-8 was added to the standard duplex ODN-1 * ODN-2. The reaction was performed at 22°C and the pyrene monomer emission increase was followed. Due to the formation of the new invader-substrate duplex ODN-1 * ODN-8, the fluorescence intensity increased by about a factor of 3 (Figure 3, left spectrum). Excitation occurred at 345 nm and emission at 398 nm. The displacement reaction was performed with a series of oligonucleotides containing stabilizer 2 in which the number of incorporations and position was altered. Plotting the Table 3. T m values of oligonucleotide duplexes containing 2,6-diamino nucleosides 1-4 or the bis-pyrene linker 5. [a] Duplexes Only T m [b] [°C] Duplexes Only T m [b] [°C] Duplexes Only T m [b] [°C] Duplexes Only T m [b] [°C] Measured at 260 nm at a concentration of 5 μM + 5 μM single strand at a heating rate of 1.0°C/min in 100 mM NaCl, 10 mM MgCl 2 , and 10 mM Na-cacodylate (pH 7.0). [b] T m values were calculated from the heating curves using the program Meltwin 3.0. [18] The standard deviation for the T m values is � 0.5°C. "Duplexes Only" means duplexes without the presence of invader or released strands.

Chemistry-A European Journal
Research Article doi.org/10.1002/chem.202202412 fluorescence change at an emission wavelength of 398 nm against the time scale reveals an exponential growth in the beginning of the displacement reaction, which ends in an asymptotic behavior after prolonged reaction time. The graphs are shown in Figure 3 together with the steady-state fluorescence curves of the starting duplex and the product duplex.
Under those conditions, displacement is fast. From the timedependent measurements, half-life values of below 100 s were calculated. The final equilibrium of the displacement reaction was observed after several minutes. The reaction progress of other displacements performed on invaders with a different number of stabilizers is similar. Displacement monitoring started about 5 s after the initiation of the reaction. The delay was the result of sample mixing before measurement and the start of the fluorescence spectrometer. From these experiments, it is obvious that the base pair stabilizer 2 can be used to drive the displacement reaction and a toehold is unnecessary.
In system A, stabilizer and sensor were implemented in the single stranded invader. In system B, the sensor is part of the target duplex (Scheme 5). System B was designed to compare the displacement reaction of oligonucleotide invaders contain-

Chemistry-A European Journal
Research Article doi.org/10.1002/chem.202202412 ing either three incorporations of stabilizer 1 or three incorporations of stabilizer 2. After displacement, only the stabilizers 1 or 2 are part of the final duplexes, and the sensor is released with the output strand. According to the bis-pyrene fluorescence behavior, excimer emission decreases after displacement. From Figure 4, it is apparent that the less efficient stabilizer 1 leads to a slower displacement as the effective stabilizer 2. This is in line with the higher thermodynamic stability of the duplexes induced by stabilizer 2 compared to that of stabilizer 1. It slows-down the reaction. Other factors can be excluded as system B differs only in the structures of stabilizers. According to the stabilizing property of the clickable , d), f) Reaction progress followed by fluorescence changes. g) Structure of the nucleoside pyrene sensor used in system A. All measurements were performed at an excitation wavelength of 345 nm with 5 μM single-strand concentration in 100 mM NaCl, 10 mM MgCl 2 , and 10 mM Na-cacodylate (pH 7.0). Scheme 5. Displacement reaction according to system B using pyrene sensor 5 and stabilizers 1 and 2.

Chemistry-A European Journal
Research Article doi.org/10.1002/chem.202202412 compound 3 and its ability to develop fluorescence after click reaction (!4), this compound comprises already the capacity to act as base pair stabilizer and as sensor in a combined system.

Conclusions
Purine-2,6-diamine (2-aminoadenine) and 8-aza-7-deaza-7-bromopurine-2,6-diamine 2'-deoxyribonucleosides 1 and 2 have been used to strengthen invader strand interactions in DNA strand displacement. They were incorporated in oligonucleotides and implemented in isothermal displacement systems. Displacement reactions were performed at 22°C without transferring energy to the system. Displacement is driven by negative enthalpy changes between the target and the displaced duplex. The displacement was performed with 12mer oligonucleotides duplexes and single strand invaders with single and multiple incorporations of modified nucleosides. T m data and thermodynamic values obtained from UV melting profiles were used to set up the displacement systems. From those, the applicability of the base stabilizers in strand displacement was established.
For strand displacement reactions, sensors were implemented that monitor the progress of the displacement. Two new environmental sensitive fluorescent pyrene sensors (4 and 5) were designed. In one sensor pyrene was connected to the nucleobase and was present in the invader strand, the other sensor contained a bis-pyrene unit and was implemented in a dendritic linker. The constructions of both sensors base on click chemistry, and phosphoramidites were prepared for solid-phase synthesis. One sensor shows monomer and the other excimer emission and both show fluorescence changes when they are part of duplexes or single strands. This fluorescence change was used to follow strand exchange in two systems; A and B. In both systems displacement is fast and efficient. Our work demonstrates that base pair stabilizers can expand the utility of displacement reactions. Toeholds as used in other displacement systems are not required. Other displacement set-ups are possible on the basis of our results in the realm of DNA and RNA. Due to modified base incorporation, invader strand and displaced duplexes might be not cleaved by nucleolytic enzymes or cleavage will be retarded. This makes the systems applicable to living matter. Applications in chemical biology, nanotechnology or for diagnostic or therapeutic uses are feasible.

Experimental Section
General: All chemicals and solvents were of laboratory grade as obtained from commercial suppliers and were used without further purification. Thin-layer chromatography (TLC) was performed on TLC aluminium sheets covered with silica gel 60 F254 (0.2 mm). Flash column chromatography (FC): silica gel 60 (40-60 μM) at 0.4 bar. UV-spectra were recorded on a UV-spectrophotometer: λ max (ɛ) in nm, ɛ in dm 3 mol À 1 cm À 1 . NMR spectra were measured at 599.74 MHz, 399.89 MHz or 300.15 MHz for 1 H, at 150.82 MHz, 100.56 MHz or 75.47 MHz for 13 C and at 121.5 MHz for 31 P. 1 H-13 C correlated (HMBC, HSQC) NMR spectra were used for the assignment of the 13 C signals (Tables S1, S2, Supporting Information). The J values are given in Hz; δ values in ppm relative to Me 4 Si as internal standard. For NMR spectra recorded in DMSO-d 6 , the chemical shift of the solvent peak was set to 2.50 ppm for 1 H NMR and 39.50 ppm for 13 C NMR. ESI-TOF mass spectra of nucleosides were recorded on a Micro-TOF spectrometer.
Oligonucleotide syntheses and characterization: Solid-phase oligonucleotide syntheses were performed with an ABI 392-08 synthesizer at 1 μmol scale (trityl-on mode) employing the phosphoramidites 8-10, as well as standard building blocks, giving an average coupling yield of over 95 %. After cleavage from the solid support, the oligonucleotides were deprotected in 28 % aqueous ammonia at 55°C for 12 h. The 4,4'-dimethoxytrityl containing oligonucleotides were purified by reversed-phase HPLC

Fluorescence Studies
General: Fluorescence measurements were performed using a F-7000 fluorescence spectrophotometer (Hitachi, Tokyo, Japan). Cleaned and dried Hellma analytics cuvettes (2 mL volume) with caps were used for the fluorescence study. All measurements were performed at ambient temperature (22°C).
Fluorescence of oligonucleotide duplexes: 5 μM + 5 μM (sensor 4) or 2.5 μM + 2.5 μM (sensor 5) of single-stranded oligonucleotides were added to 1 mL buffer (100 mM NaCl, 10 mM MgCl 2 , and 10 mM Na-cacodylate, pH 7.0). To determine the excitation wavelengths of the sensors, UV spectra of oligonucleotides were measured. The long wavelengths UV maximum were used for the determination of the emission of the oligonucleotide. The monochromator was set to the long wavelength maximum of fluorescence emission, and the wavelength of the excitation monochromator is scanned across the desired excitation range. Then, the intensity of fluorescence is recorded on the detector as a function of excitation wavelength. The long-wavelength maximum from the excitation spectrum was used to measure the emission spectra. Accordingly, oligonucleotides containing sensor 4 were excited at 345 nm, oligonucleotides incorporating sensor 5 were excited at 347 nm. Oligonucleotides with sensor 4 show monomer emission around 384 and 398 nm, oligonucleotides with the dendritic bis-pyrene linker 5 show excimer emission at 488 nm. The UV absorbance was below 0.01 to exclude inner filter effects.
Reaction progress of the strand displacement reactions: To a solution containing 5 μM (sensor 4) or 2.5 μM (sensor 5) of the particular duplex in a fluorescence cuvette, the invader strand was added in the same concentration by a micro pipette. Then, the cuvette was shaken vigorously and set in the spectrophotometer. Then, time dependent fluorescence measurements were carried out at room temperature for a particular time interval. The instrument parameters were set as follows: Excitation wavelength (EX) 345 nm, emission wavelength (EM) 398 nm for system A, EX 347 nm, EM 488 nm for system B. Slit widths were set to 10 nm for excitation and 10 nm for emission. Fluorescence data were processed using the program OriginPro 2017.