Phosphoramidite building blocks with protected nitroxides for the synthesis of spin-labeled DNA and RNA

TEMPO spin labels protected with 2-nitrobenzyloxymethyl groups were attached to the amino residues of three different nucleosides: deoxycytidine, deoxyadenosine, and adenosine. The corresponding phosphoramidites could be incorporated by unmodified standard procedures into four different self-complementary DNA and two RNA oligonucleotides. After photochemical removal of the protective group, elimination of formic aldehyde and spontaneous air oxidation, the nitroxide radicals were regenerated in high yield. The resulting spin-labeled palindromic duplexes could be directly investigated by PELDOR spectroscopy without further purification steps. Spin–spin distances measured by PELDOR correspond well to the values obtained from molecular models.

Subsequently conc. NaHCO 3 solution was added and the mixture was extracted with CH 2 Cl 2 . The combined organic layers were dried with MgSO 4 and the solvent was evaporated under reduced pressure. After purification by silica gel chromatography (CH 2 Cl 2 /MeOH 9:1) 15 could be obtained as a light yellow foam (2.45 g, 90%). R f = 0.46 (CH 2 Cl 2 /MeOH 9:1). 1
After the reaction mixture was allowed to warm up to ambient temperature, it was stirred for 23 h and cooled down to 0 °C again. The reaction was quenched with MeOH and stirred for 15 min at 0 °C.      1.00 equiv) and Et 3 N (0.4 mL, 2.68 mmol, 5.00 equiv) in 20 mL CH 2 Cl 2 N,N-diisopropylamino(2cyanoethyl)phosphoramidic chloride (0.25 g, 1.07 mmol, 2.00 equiv) was added dropwise. After stirring for 23 h at ambient temperature conc. NaHCO 3 solution was added, the organic layer was separated and the aqueous phase was extracted with CH 2 Cl 2 . The combined organic layers were dried with MgSO 4 and the solvent was removed under reduced pressure. Purification by silica gel chromatography (CH 2 Cl 2 /EtOAc/Et 3 N 80:20:1) gave amidite 8 as a colourless foam (0.54 g, 86%). R f S10 = 0.85, 0.66 (CH 2 Cl 2 /EtOAc 4:1 (mixture of 2 diastereomers)). 1     The column was heated to 55 °C in most cases (55 °C and 20 °C for 27c).
Quantification. Oligonucleotide concentrations were determined via UV spectrometry on a nanodrop2000 (Thermo Scientific) using Lambert-Beer´s law. Extinction coefficients were calculated by a nearest neighbor model according to literature [5]. For modified bases identical increments were used as for their natural counterparts.      and subsequently annealed (see Table S1). Conditions for RP-HPLC:   Table S1 below. This reaction is shown as an example. Insets: LC-MS analysis of hemiacetal 26b and of spin-labeled RNA 26c.

Hemiacetal 26b
Radical 26c      After preparative separation, peak 1 and peak 2 form a mixture of both after standing or induced by a second annealing procedure ( Figure S18). peak 1 peak 2 Figure S18: Overlay of peak 1 (black) and 2 (red) after clean separation and a second annealing step. The chromatograms show a conversion of peak 1 into peak 2 and vice versa suggesting a conformational equilibrium. If isolated peak 1 or 2 is kept at room temperature for several days, in both cases peak 1 dominates by far. A cautious interpretation is that the second peak might correspond to a stem-loop structure and peak 1 to the duplex, in accordance with PELDOR data ( Figure S21).

S21
EPR method part cw-EPR before and after annealing. Continuous wave (cw) EPR spectra were measured at X-band The presence of monomeric species containing single spin labels is also visible in lower levels of modulation depth ( Figure S21) for all DNA samples, in particular for 22c and 24c.

S22
increased after the following step of annealing, which leads to mean spin labeling efficiencies around 96%.
PELDOR distance measurements. 10 [7] was applied with pulse lengths of 22 ns for the detection pulses (π/2 and π) and 12 ns for the pump pulse (π). The pump pulse frequency was set to the maximum of the echo detected field sweep spectrum and the frequency of the detection pulses 70 MHz lower. The first interpulse delay was increased by 16 ns for eight steps Figure S20: cw-EPR spectra of 22c-27c a) directly after irradiation (grey) and b) after additional annealing (black).

S23
to avoid deuterium modulation. Figure S21 shows the results of the PELDOR measurements for the samples 22c-27c and Table S2 compares the experimentally obtained distances with the simulations. Figure S21: PELDOR measurements of 22c-27c. After correction of the intermolecular exponential background (red) from the time traces V(t)/V(0) the form factors F(t)/F(0) were obtained. They were fitted with a model-free Tikhonov regularization (DeerAnalysis15) [8]. On the form factors the fits are superimposed (red). The distance distributions P(r) show distinct values and the asterisks indicate additional distances for the RNA sample 27c, probably due to stacking of the RNA. DNA samples show reduced levels of modulation depth caused by the presence of monomeric strands ( Figure S19). For example, the modulation depths λ of the palindromic 12mers can be compared with λ of an ideal 2-spin model system (λ = 0.31) to estimate the amount of the duplex structure. Taking the spin labeling efficiencies into account, the modulation depths suggest 100% duplex structure for sample 26c (λ = 0.31) and roughly 45% and 25% duplex structure for samples 22c (λ = 0.13) and 24c (λ = 0.08), respectively. The distances predicted by molecular modeling are shown in blue.

Spin-spin-distances in palindromic duplexes
Simulation of the spin-spin distances. All deoxyribonucleic acids (22c-25c) were generated as a Bform duplex using SPARTAN [9]. Ribonucleic acids (26c, 27c) were built as an A-form duplex. The attachment of the spin label was done with SPARTAN as well. After that, a local optimization, based on the force field MMFF94, was carried out applying AVOGADRO [10]. Optimization was executed twice for each duplex.

Deoxyadenosine phosphoramidite with protected spin label (7)
This report was created by ACD/NMR Processor Academic Edition. For more information go to www.acdlabs.com/nmrproc/