Supporting information Investigating Functional DNA Grafted on ...
Supporting information
Investigating Functional DNA Grafted on Nanodiamond Surface Using Site-Directed Spin Labeling and Electron Paramagnetic Resonance Spectroscopy
Rana D. Akiel,1,3 Xiaojun Zhang,1 Chathuranga Abeywardana,1,3 Viktor Stepanov,1,3 Peter Z. Qin,1 and Susumu Takahashi1,2,3 1
Department of Chemistry, University of Southern California, Los Angeles CA 90089,
USA 2
Department of Physics, University of Southern California, Los Angeles CA 90089, USA
3
Center for Quantum Information Science & Technology (CQIST), University of
Southern California, Los Angeles CA 90089, USA
S1
Table of Contents S.1.
Synthesis and HPLC characterization of DBCO-tagged DNA (S1-DBCO) ................. S3
S.2.
Cu-free click reaction to attach DBCO on ND ............................................................ S3
S.3.
Cu-free click reaction to attach DBCO-tagged DNA (S1-DBCO) to ND ...................... S4
S.4.
Characterization of Cu-free click reaction in solution by EPR spectroscopy ............... S4
S.5.
Determination of the yield of click reactions ............................................................... S5
S.6. Verification of the wash procedure for removing DNA strands non-specifically associated with ND ............................................................................................................... S6 S.7.
Hybridization of S1-ND with the complimentary strand (S2) ....................................... S7
S.8.
Simulation of EPR spectra obtained from nitroxide-labeled S2-S1-ND ...................... S8
S.9. Similarity between EPR spectra measured from R5a-labeled DNA duplexes tethered to ND and Streptavidin ........................................................................................................... S10 S.10. Synthesis of TEMPO-ND and characterization by X-band and 115 GHz EPR spectroscopy ...................................................................................................................... S11 References ............................................................................................................................ S12
S2
S.1.
Synthesis and HPLC characterization of DBCO-tagged DNA (S1-DBCO)
S1-DBCO was synthesized following Scheme S1. After the reaction, the product was purified using anion-exchange high performance liquid chromatography (AX HPLC) followed by reverse-phase HPLC (RP HPLC). Figure S2 shows the HPLC traces of the DBCO-NHS, DNA S1-NH2 precursor, and the S1-DBCO product. The reaction product (S1-DBCO) clearly elutes later than either the DNA or the DBCO, indicating successful coupling.
Scheme S1. Synthesis of S1-DBCO.
Figure S2. Examples of anion-exchange HPLC traces showing precursors (top: DBCO-NHS; middle: DNA S1-NH2) and the DBCO-reacted product (bottom).
S.2.
Cu-free click reaction to attach DBCO on ND
Scheme S3 shows the Cu-free click reaction scheme to attach DBCO on the ND surface.
Scheme S3. Copper-free click reaction of DBCO and ND-N3. S3
S.3.
Cu-free click reaction to attach DBCO-tagged DNA (S1-DBCO) to ND
Scheme S4 shows the Cu-free click reaction scheme to attach DBCO-tagged DNA on the ND surface.
Scheme S4. Copper-free click reaction of DBCO-S1 and ND-N3.
S.4.
Characterization of Cu-free click reaction in solution by EPR spectroscopy
To characterize the applicability of the copper-free click reaction, DBCO-S1 was reacted with a nitroxide precursor, 4 azido-2,2,6,6-Tetramethyl-1-piperidinyloxy (TEMPO-N3, kindly provided by Dr. Kálmán Hideg, University of Pécs, Hungary) (Scheme S5). The reaction was carried out in a mixture (50 µL) with approximately 20 µM of S1-DBCO, 50 µM of TEMPO-N3, and 50%/50% (v/v) acetonitrile/water. The solution was incubated for 22 hours at room temperature with constant mixing. The resulting product was purified with HPLC.
Scheme S5. Copper-free click reaction of DBCO-S1 and TEMPO-N3. Measured EPR spectra of the labeled DNA showed the characteristic three-peak pattern from 14N nitroxide radicals (Figure S6), with the nitroxide rotational correlation time (τc) determined to be 0.57±0.04 ns. The obtained correlation time is slower than free nitroxide radicals (Figure S6 inset), and similar to the reported value for spin-labeled single-stranded DNAs.1 Therefore, the result confirmed successful copper-free click reaction with S1-DBCO.
S4
Figure S6. EPR spectrum of S1-TEMPO. The correlation time (τc) was determined to be 0.57±0.04 ns based on EPR lineshape simulation (solid blue line). The inset shows EPR data of a free nitroxide, which gives τc = 0.03±0.01 ns under the same experimental conditions. Lineshape simulations were performed using Easyspin2 with the isotropic motion model and the TEMPO spin Hamiltonian (S = 1/2, gx = 2.0085, gy = 2.0059, gz = 2.0021, Ax = 6.5 MHz, Ay = 5.6 MHz, Az = 37 MHz).
S.5.
Determination of the yield of click reactions
The yield of click reactions was determined by analyzing the reduction in FTIR peak intensity at 2100 cm-1, which represents the N3 signal. Note that FTIR intensity depends on the amount of ND, which is proportional to the amount of N3 groups. Therefore, we first calibrated the FTIR intensity as a function of the ND weight. Figure S7 shows a plot of averaged intensity at 2100 cm-1 vs. ND weight, and a linear dependence can be clearly observed. In addition, variations at each data points were small, indicating a homogeneous distribution of the ND samples in the pellets being measured. This calibration study allowed us to define the normalized FTIR intensity for unreacted NDs (I0) as: I0= (1.9±0.2) × (the weight of NDs in mg).
(1)
Then, using the measured intensity from the weight-normalized FTIR spectra (Iexp) and the normalized intensity (Equation 1), the reaction yield, defined as the percentage of azide groups reacted, is then calculated as: Yield (%) = 100 × (I0 - Iexp)/I0.
(2)
Figure S8 shows representative ND-weight-normalized FTIR spectra at 2100 cm-1. The normalization was done using Equation 1. FTIR signal was clearly reduced for N3-ND reacted with S1-DBCO, but remained unchanged with S1 (i.e., N3-ND mixed with S1 in the absence of DBCO). As shown in Figure 1 and Figure S8, the click reaction yield can be driven to close to 100% if sufficient amount of S1-DBCO can be supplied in the reaction.
S5
Figure S7. FTIR intensity of the N3 signal (at 2100 cm-1) as a function of ND weight. The solid line is a fit to a linear function. The measurements were performed five times at each sample. The mean and distribution in the five measurements are represented by the square point and error bar.
Figure S8. Representative ND-weight-normalized FTIR signal at the 2100 cm-1 region.
S.6. Verification of the wash procedure for removing DNA strands nonspecifically associated with ND The wash procedure was verified by monitoring EPR signals from a spin-labeled DNA strand, R5(p4)-S2, mixed with N3-ND (i.e., without the S1 strand attached).
S6
First, samples with known R5(p4)-S2 concentrations in solution (without exposed to ND) were measured, and the resulting EPR spectra were analyzed using Easyspin2 to obtain the respective intensity. Plotting the EPR intensity vs. DNA concentration yielded the calibration curved shown in Figure S9(a). R5(p4)-S2 was then incubated with N3-ND (i.e., without the S1 strand attached), and the mixture was subjected to wash by either 80%/20% (v/v) acetonitrile/water or PBS as described in the Methods section. Following each wash, EPR spectrum of the supernatant was measured using the same set of acquisition parameters as that for determining the calibration curve; the intensity of the EPR spectrum was obtained using Easyspin analysis; and the corresponding DNA concentration was determined based on the calibration curve. As shown in the inset of Figure S9(b), the EPR intensity decreased significantly after wash. Figure S9(b) shows the amount of DNA present in the supernatant as a function of the number of wash. Using either 80%/20% (v/v) acetonitrile/water or PBS solutions, the amount of DNA in the supernatant quickly reduced and became too small to measure after only a few washes. The data also indicated that the acetonitrile solution is more effective in removing the nonspecifically associated DNA. Furthermore, upon conclusion of the wash process, we measured EPR spectra of the precipitates. As shown in Figure S9(c), only EPR signals of NDs were observed in the precipitates, again confirming that all the R5(p4)-S2 has been removed. Overall, the experiments indicate that non-specifically associated DNA strands can be effectively removed from ND using the wash procedure described in the main text.
Figure S9. (a) EPR intensity as a function of the known concentration of R5(p4)-S2 in PBS. The EPR intensity analyses were performed wtih S = 1/2, gx = 2.0085, gy = 2.0059, gz = 2.0021, Ax = 6.5 MHz, Ay =5.6 MHz and Az = 37 MHz, 0.01 mT of the linewidth and 0.4 ns of the rotational correlation time were used. (b) A plot of the DNA concentration in the supernatant vs. the number of wash. Inset shows the corresponding EPR spectra obtained from washes using PBS solution, which were measured with identical acquisition parameters and were not subjected to spectral normalization. The EPR intensity analyses were performed with 0.4 ns and 0.6 ns of the rotational correlation time for the PBS and acetonitrile solutions, respectively. (c) EPR spectra of the precipitates after the wash process. The spectra was simulated using spin parameters of NDs:3 14N impurity: S = 1/2, gx,y = 2.0024, gz = 2.0024, Ax,y = 82 MHz, and Az = 114 MHz; surface impurities: S = 1/2, gx,y = 2.0029 and gz = 2.0027.
S.7.
Hybridization of S1-ND with the complimentary strand (S2)
Scheme S10 shows sequences of the S1 strand and the S2 strand, which is complementary to the S1 strand. S7
Scheme S10. S2-S1 duplex tethered to ND. Arrows indicate the backbone phosphate sites at the S2 strand at which spin labels have been attached.
S.8.
Simulation of EPR spectra obtained from nitroxide-labeled S2-S1-ND
We analyzed the EPR spectra using Easyspin2 with known spin parameters. For the ND component3: 14N impurity: S = 1/2, gx,y = 2.0024, gz = 2.0024, Ax,y = 82 MHz, and Az = 114 MHz; surface impurities: S = 1/2, gx,y = 2.0029 and gz = 2.0027. For the nitroxide component: S = 1/2, gx = 2.0085, gy = 2.0059, gz = 2.0021, Ax = 6.5 MHz, Ay =5.6 MHz and Az = 37 MHz. In the simulations, the linewidth was set at 0.01 mT. In the simulation of Easyspin, the solid-state package (pepper) was used for ND while an isotropic motion with the slow-motion package (chili) was used for nitroxide. Figures S11–S14 show the simulation results, with blue solid line: experimental data; red dotted line: sum of the simulated ND and nitroxide components; green solid line: simulated ND component; light blue solid line: the subtracted experimental spectrum by the ND component; magenta solid line: simulated nitroxide component.
Figure S11. Spectral simulation of (p4)-S2-S1-ND (1st hybridization). The nitroxide component was simulated using an isotropic motion model, and the simulation gave a rotational correlation time of 1.0 ns. S8
Figure S12. Spectral simulation of R5(p4)-S2-S1-ND (2nd hybridization). The nitroxide component was simulated using an isotropic motion model, and the simulation gave a rotational correlation time of 1.0 ns.
Figure S13. Spectral simulation of R5(p17)-S2-S1-ND. The nitroxide component was simulated using an isotropic motion model, and the simulation gave a rotational correlation time of 1.2 ns.
S9
Figure S14. Spectral simulation of R5a(p4)-S2-S1-ND. The nitroxide simulation was carried out using the microscopic ordered macroscopic disordered (MOMD) model, with Rx = 107.95 s-1, Ry = 108.42 s-1, Rz = 107.8 s-1, αD = 2.01º, βD = 35º, γD = 60.62º, and the orienting potential coefficients (λ20) to be 1.3. The corresponding rotational correlation time (τc = 1/(6Rave) where Rave = (Rx+Ry+Rz)/3) was found to be 1.2 ns.
S.9. Similarity between EPR spectra measured from R5a-labeled DNA duplexes tethered to ND and Streptavidin Figure S15 shows EPR spectra of R5a(p4)-S2-S1-Streptadidine and R5a(p4)-S2-S1-ND. Both spectra show splitting at the low-field peak (indicated by dotted circles), which arise from anisotropic rotation under a restricted potential.4
Figure S15. X-band EPR spectra of R5a(p4)-S2-S1 tethered to Streptavidin (left) and ND (right). The Streptavidin spectrum was reproduced from ref.(5), and the ND spectrum was the nitroxide component shown in Figure S14. S10
S.10. Synthesis of TEMPO-ND and characterization by X-band and 115 GHz EPR spectroscopy Scheme S16 shows the procedure for coupling a nitroxide to ND using copper-free click reaction. The nitroxide radical used in this reaction is 4-Amino-2,2,6,6-tetramethylpiperidine-1oxyl (4-amino-TEMPO, Sigma-Aldrich, Milwaukee, #163945). In the first step, the reaction was carried out with 1 eq. of DBCO and 1.5 eq. of 4-amino-TEMPO. After incubation overnight, the reaction solution was evaporated using a speedvac, then the mixture was suspended in 80%/20% (v/v) acetonitrile/water. The mixture was reacted without further purification with N3ND for 22 hours at room temperature under constant ultrasonication. Then excess reactants were washed away using multiple washing/centrifugation cycles in acetonitrile until no EPR signal of TEMPO was observed in the supernatant.
Scheme S16. Synthesis of TEMPO-ND. 115 GHz EPR spectroscopy was performed using a home-built EPR spectrometer. The 115 GHz EPR system employs a high-power solid-state source consisting 8–10 GHz synthesizer, pi-n switch, microwave amplifiers, and frequency multipliers. The output power of the source system is 700 mW at 115 GHz. 115 GHz excitation is propagated using quasioptical bridge and a corrugated waveguide and couples to a sample located at the center of 12.1 T cryogenic-free superconducting magnet. EPR signals are isolated from the excitation using an induction mode operation. For EPR detection, we employed a superheterodyne detection system in which 115 GHz is down-converted into 3 GHz of intermediate frequency (IF), then down-converted again to in-phase and quadrature components of dc signals. Details of the system have been described elsewhere.6, 7 To obtain a cw EPR spectrum, the sample was placed in a Teflon sample holder (5 mm diameter), with a typical volume being 1 µL. Microwave power and magnetic field modulation strength were adjusted to maximize the intensity of EPR signals without distorting the lineshape (typically, the microwave power of 2 mW, modulation amplitude of 0.1 mT with modulation frequency of 20 kHz). The measured spectra were shown in Figure S17. These spectra were analyzed using Easyspin.2 The parameters are: (i) for the ND component, 14N impurity: S = 1/2, gx,y = 2.0024, gz = 2.0024, Ax,y = 82 MHz, and Az = 114 MHz; surface impurities: S = 1/2, gx,y = 2.0029 and gz = 2.0027; and (ii) for the nitroxide component: S = 1/2, gx = 2.0086, gy = 2.0056, gz = 2.0033, Ax = 6.5, Ay = 5.6, Az = 37 MHz; and spectral linewidth = 0.5 mT. By analyzing the best-fit spectra obtained from simulations at both the X-band (9.3 GHz) and 115 GHz, the rotational correlation time of the ND-tethered TEMPO (τc) was determined to be 4.1 ns.
S11
Figure S17. EPR spectra of TEMPO-ND. Blue solid lines represent the experimental data, red dotted lines are the simulated spectra, green solid lines are ND components, and magenta solid lines represent the nitroxide components. (a) X-band (~9.3 GHz) spectrum. The nitroxide signal was magnified by 2 times to show the EPR lineshape. (b) 115 GHz EPR spectrum. The nitroxide signal was magnified by 10 times to show the EPR lineshape.
References 1. Grant, G. P. G. and Qin, P. Z., A facile method for attaching nitroxide spin labels at the 50 terminus of nucleic acids. Nucl. Acids Res. 2007, 35, 1-8. 2. Stoll, S. and Schweiger, A., EasySpin, a comprehensive software package for spectral simulation and analysis in EPR. J. Magn. Reson. 2006, 178, 42-55. 3. Romanova, E. E., Akiel, R., Cho, F. H. and Takahashi, S., Grafting nitroxide radicals on nanodiamond surface using click chemistry. J. Phys. Chem. A 2013, 117, 11933–11939. 4. Popova, A. M., Kálai, T., Hideg, K. and Qin, P. Z., Site-specific DNA structural and dynamic features revealed by nucleotide-independent nitroxide probes. Biochemistry 2009, 48, 85408550. 5. Ding, Y., Zhang, X., Tham, K. W. and Qin, P. Z., Experimental mapping of DNA duplex shape enabled by global lineshape analyses of a nucleotide-independent nitroxide probe. Nucl. Acids Res. 2014, 42, e140. 6. Cho, F. H., Stepanov, V., Abeywardana, C. and Takahashi, S., 230/115 GHz electron paramagnetic resonance/double electron-electron resonance spectroscopy. Methods Enzymol. 2015, 563, 95-118. 7. Cho, F. H., Stepanov, V. and Takahashi, S., A high-frequency electron paramagnetic resonance spectrometer for multi-dimensional, -frequency and -phase pulsed measurements. Rev. Sci. Instrum. 2014, 85, 075110.
S12
Investigating Functional DNA Grafted on Nanodiamond Surface Using Site-Directed Spin Labeling and Electron Paramagnetic Resonance Spectroscopy
Rana D. Akiel,1,3 Xiaojun Zhang,1 Chathuranga Abeywardana,1,3 Viktor Stepanov,1,3 Peter Z. Qin,1 and Susumu Takahashi1,2,3 1
Department of Chemistry, University of Southern California, Los Angeles CA 90089,
USA 2
Department of Physics, University of Southern California, Los Angeles CA 90089, USA
3
Center for Quantum Information Science & Technology (CQIST), University of
Southern California, Los Angeles CA 90089, USA
S1
Table of Contents S.1.
Synthesis and HPLC characterization of DBCO-tagged DNA (S1-DBCO) ................. S3
S.2.
Cu-free click reaction to attach DBCO on ND ............................................................ S3
S.3.
Cu-free click reaction to attach DBCO-tagged DNA (S1-DBCO) to ND ...................... S4
S.4.
Characterization of Cu-free click reaction in solution by EPR spectroscopy ............... S4
S.5.
Determination of the yield of click reactions ............................................................... S5
S.6. Verification of the wash procedure for removing DNA strands non-specifically associated with ND ............................................................................................................... S6 S.7.
Hybridization of S1-ND with the complimentary strand (S2) ....................................... S7
S.8.
Simulation of EPR spectra obtained from nitroxide-labeled S2-S1-ND ...................... S8
S.9. Similarity between EPR spectra measured from R5a-labeled DNA duplexes tethered to ND and Streptavidin ........................................................................................................... S10 S.10. Synthesis of TEMPO-ND and characterization by X-band and 115 GHz EPR spectroscopy ...................................................................................................................... S11 References ............................................................................................................................ S12
S2
S.1.
Synthesis and HPLC characterization of DBCO-tagged DNA (S1-DBCO)
S1-DBCO was synthesized following Scheme S1. After the reaction, the product was purified using anion-exchange high performance liquid chromatography (AX HPLC) followed by reverse-phase HPLC (RP HPLC). Figure S2 shows the HPLC traces of the DBCO-NHS, DNA S1-NH2 precursor, and the S1-DBCO product. The reaction product (S1-DBCO) clearly elutes later than either the DNA or the DBCO, indicating successful coupling.
Scheme S1. Synthesis of S1-DBCO.
Figure S2. Examples of anion-exchange HPLC traces showing precursors (top: DBCO-NHS; middle: DNA S1-NH2) and the DBCO-reacted product (bottom).
S.2.
Cu-free click reaction to attach DBCO on ND
Scheme S3 shows the Cu-free click reaction scheme to attach DBCO on the ND surface.
Scheme S3. Copper-free click reaction of DBCO and ND-N3. S3
S.3.
Cu-free click reaction to attach DBCO-tagged DNA (S1-DBCO) to ND
Scheme S4 shows the Cu-free click reaction scheme to attach DBCO-tagged DNA on the ND surface.
Scheme S4. Copper-free click reaction of DBCO-S1 and ND-N3.
S.4.
Characterization of Cu-free click reaction in solution by EPR spectroscopy
To characterize the applicability of the copper-free click reaction, DBCO-S1 was reacted with a nitroxide precursor, 4 azido-2,2,6,6-Tetramethyl-1-piperidinyloxy (TEMPO-N3, kindly provided by Dr. Kálmán Hideg, University of Pécs, Hungary) (Scheme S5). The reaction was carried out in a mixture (50 µL) with approximately 20 µM of S1-DBCO, 50 µM of TEMPO-N3, and 50%/50% (v/v) acetonitrile/water. The solution was incubated for 22 hours at room temperature with constant mixing. The resulting product was purified with HPLC.
Scheme S5. Copper-free click reaction of DBCO-S1 and TEMPO-N3. Measured EPR spectra of the labeled DNA showed the characteristic three-peak pattern from 14N nitroxide radicals (Figure S6), with the nitroxide rotational correlation time (τc) determined to be 0.57±0.04 ns. The obtained correlation time is slower than free nitroxide radicals (Figure S6 inset), and similar to the reported value for spin-labeled single-stranded DNAs.1 Therefore, the result confirmed successful copper-free click reaction with S1-DBCO.
S4
Figure S6. EPR spectrum of S1-TEMPO. The correlation time (τc) was determined to be 0.57±0.04 ns based on EPR lineshape simulation (solid blue line). The inset shows EPR data of a free nitroxide, which gives τc = 0.03±0.01 ns under the same experimental conditions. Lineshape simulations were performed using Easyspin2 with the isotropic motion model and the TEMPO spin Hamiltonian (S = 1/2, gx = 2.0085, gy = 2.0059, gz = 2.0021, Ax = 6.5 MHz, Ay = 5.6 MHz, Az = 37 MHz).
S.5.
Determination of the yield of click reactions
The yield of click reactions was determined by analyzing the reduction in FTIR peak intensity at 2100 cm-1, which represents the N3 signal. Note that FTIR intensity depends on the amount of ND, which is proportional to the amount of N3 groups. Therefore, we first calibrated the FTIR intensity as a function of the ND weight. Figure S7 shows a plot of averaged intensity at 2100 cm-1 vs. ND weight, and a linear dependence can be clearly observed. In addition, variations at each data points were small, indicating a homogeneous distribution of the ND samples in the pellets being measured. This calibration study allowed us to define the normalized FTIR intensity for unreacted NDs (I0) as: I0= (1.9±0.2) × (the weight of NDs in mg).
(1)
Then, using the measured intensity from the weight-normalized FTIR spectra (Iexp) and the normalized intensity (Equation 1), the reaction yield, defined as the percentage of azide groups reacted, is then calculated as: Yield (%) = 100 × (I0 - Iexp)/I0.
(2)
Figure S8 shows representative ND-weight-normalized FTIR spectra at 2100 cm-1. The normalization was done using Equation 1. FTIR signal was clearly reduced for N3-ND reacted with S1-DBCO, but remained unchanged with S1 (i.e., N3-ND mixed with S1 in the absence of DBCO). As shown in Figure 1 and Figure S8, the click reaction yield can be driven to close to 100% if sufficient amount of S1-DBCO can be supplied in the reaction.
S5
Figure S7. FTIR intensity of the N3 signal (at 2100 cm-1) as a function of ND weight. The solid line is a fit to a linear function. The measurements were performed five times at each sample. The mean and distribution in the five measurements are represented by the square point and error bar.
Figure S8. Representative ND-weight-normalized FTIR signal at the 2100 cm-1 region.
S.6. Verification of the wash procedure for removing DNA strands nonspecifically associated with ND The wash procedure was verified by monitoring EPR signals from a spin-labeled DNA strand, R5(p4)-S2, mixed with N3-ND (i.e., without the S1 strand attached).
S6
First, samples with known R5(p4)-S2 concentrations in solution (without exposed to ND) were measured, and the resulting EPR spectra were analyzed using Easyspin2 to obtain the respective intensity. Plotting the EPR intensity vs. DNA concentration yielded the calibration curved shown in Figure S9(a). R5(p4)-S2 was then incubated with N3-ND (i.e., without the S1 strand attached), and the mixture was subjected to wash by either 80%/20% (v/v) acetonitrile/water or PBS as described in the Methods section. Following each wash, EPR spectrum of the supernatant was measured using the same set of acquisition parameters as that for determining the calibration curve; the intensity of the EPR spectrum was obtained using Easyspin analysis; and the corresponding DNA concentration was determined based on the calibration curve. As shown in the inset of Figure S9(b), the EPR intensity decreased significantly after wash. Figure S9(b) shows the amount of DNA present in the supernatant as a function of the number of wash. Using either 80%/20% (v/v) acetonitrile/water or PBS solutions, the amount of DNA in the supernatant quickly reduced and became too small to measure after only a few washes. The data also indicated that the acetonitrile solution is more effective in removing the nonspecifically associated DNA. Furthermore, upon conclusion of the wash process, we measured EPR spectra of the precipitates. As shown in Figure S9(c), only EPR signals of NDs were observed in the precipitates, again confirming that all the R5(p4)-S2 has been removed. Overall, the experiments indicate that non-specifically associated DNA strands can be effectively removed from ND using the wash procedure described in the main text.
Figure S9. (a) EPR intensity as a function of the known concentration of R5(p4)-S2 in PBS. The EPR intensity analyses were performed wtih S = 1/2, gx = 2.0085, gy = 2.0059, gz = 2.0021, Ax = 6.5 MHz, Ay =5.6 MHz and Az = 37 MHz, 0.01 mT of the linewidth and 0.4 ns of the rotational correlation time were used. (b) A plot of the DNA concentration in the supernatant vs. the number of wash. Inset shows the corresponding EPR spectra obtained from washes using PBS solution, which were measured with identical acquisition parameters and were not subjected to spectral normalization. The EPR intensity analyses were performed with 0.4 ns and 0.6 ns of the rotational correlation time for the PBS and acetonitrile solutions, respectively. (c) EPR spectra of the precipitates after the wash process. The spectra was simulated using spin parameters of NDs:3 14N impurity: S = 1/2, gx,y = 2.0024, gz = 2.0024, Ax,y = 82 MHz, and Az = 114 MHz; surface impurities: S = 1/2, gx,y = 2.0029 and gz = 2.0027.
S.7.
Hybridization of S1-ND with the complimentary strand (S2)
Scheme S10 shows sequences of the S1 strand and the S2 strand, which is complementary to the S1 strand. S7
Scheme S10. S2-S1 duplex tethered to ND. Arrows indicate the backbone phosphate sites at the S2 strand at which spin labels have been attached.
S.8.
Simulation of EPR spectra obtained from nitroxide-labeled S2-S1-ND
We analyzed the EPR spectra using Easyspin2 with known spin parameters. For the ND component3: 14N impurity: S = 1/2, gx,y = 2.0024, gz = 2.0024, Ax,y = 82 MHz, and Az = 114 MHz; surface impurities: S = 1/2, gx,y = 2.0029 and gz = 2.0027. For the nitroxide component: S = 1/2, gx = 2.0085, gy = 2.0059, gz = 2.0021, Ax = 6.5 MHz, Ay =5.6 MHz and Az = 37 MHz. In the simulations, the linewidth was set at 0.01 mT. In the simulation of Easyspin, the solid-state package (pepper) was used for ND while an isotropic motion with the slow-motion package (chili) was used for nitroxide. Figures S11–S14 show the simulation results, with blue solid line: experimental data; red dotted line: sum of the simulated ND and nitroxide components; green solid line: simulated ND component; light blue solid line: the subtracted experimental spectrum by the ND component; magenta solid line: simulated nitroxide component.
Figure S11. Spectral simulation of (p4)-S2-S1-ND (1st hybridization). The nitroxide component was simulated using an isotropic motion model, and the simulation gave a rotational correlation time of 1.0 ns. S8
Figure S12. Spectral simulation of R5(p4)-S2-S1-ND (2nd hybridization). The nitroxide component was simulated using an isotropic motion model, and the simulation gave a rotational correlation time of 1.0 ns.
Figure S13. Spectral simulation of R5(p17)-S2-S1-ND. The nitroxide component was simulated using an isotropic motion model, and the simulation gave a rotational correlation time of 1.2 ns.
S9
Figure S14. Spectral simulation of R5a(p4)-S2-S1-ND. The nitroxide simulation was carried out using the microscopic ordered macroscopic disordered (MOMD) model, with Rx = 107.95 s-1, Ry = 108.42 s-1, Rz = 107.8 s-1, αD = 2.01º, βD = 35º, γD = 60.62º, and the orienting potential coefficients (λ20) to be 1.3. The corresponding rotational correlation time (τc = 1/(6Rave) where Rave = (Rx+Ry+Rz)/3) was found to be 1.2 ns.
S.9. Similarity between EPR spectra measured from R5a-labeled DNA duplexes tethered to ND and Streptavidin Figure S15 shows EPR spectra of R5a(p4)-S2-S1-Streptadidine and R5a(p4)-S2-S1-ND. Both spectra show splitting at the low-field peak (indicated by dotted circles), which arise from anisotropic rotation under a restricted potential.4
Figure S15. X-band EPR spectra of R5a(p4)-S2-S1 tethered to Streptavidin (left) and ND (right). The Streptavidin spectrum was reproduced from ref.(5), and the ND spectrum was the nitroxide component shown in Figure S14. S10
S.10. Synthesis of TEMPO-ND and characterization by X-band and 115 GHz EPR spectroscopy Scheme S16 shows the procedure for coupling a nitroxide to ND using copper-free click reaction. The nitroxide radical used in this reaction is 4-Amino-2,2,6,6-tetramethylpiperidine-1oxyl (4-amino-TEMPO, Sigma-Aldrich, Milwaukee, #163945). In the first step, the reaction was carried out with 1 eq. of DBCO and 1.5 eq. of 4-amino-TEMPO. After incubation overnight, the reaction solution was evaporated using a speedvac, then the mixture was suspended in 80%/20% (v/v) acetonitrile/water. The mixture was reacted without further purification with N3ND for 22 hours at room temperature under constant ultrasonication. Then excess reactants were washed away using multiple washing/centrifugation cycles in acetonitrile until no EPR signal of TEMPO was observed in the supernatant.
Scheme S16. Synthesis of TEMPO-ND. 115 GHz EPR spectroscopy was performed using a home-built EPR spectrometer. The 115 GHz EPR system employs a high-power solid-state source consisting 8–10 GHz synthesizer, pi-n switch, microwave amplifiers, and frequency multipliers. The output power of the source system is 700 mW at 115 GHz. 115 GHz excitation is propagated using quasioptical bridge and a corrugated waveguide and couples to a sample located at the center of 12.1 T cryogenic-free superconducting magnet. EPR signals are isolated from the excitation using an induction mode operation. For EPR detection, we employed a superheterodyne detection system in which 115 GHz is down-converted into 3 GHz of intermediate frequency (IF), then down-converted again to in-phase and quadrature components of dc signals. Details of the system have been described elsewhere.6, 7 To obtain a cw EPR spectrum, the sample was placed in a Teflon sample holder (5 mm diameter), with a typical volume being 1 µL. Microwave power and magnetic field modulation strength were adjusted to maximize the intensity of EPR signals without distorting the lineshape (typically, the microwave power of 2 mW, modulation amplitude of 0.1 mT with modulation frequency of 20 kHz). The measured spectra were shown in Figure S17. These spectra were analyzed using Easyspin.2 The parameters are: (i) for the ND component, 14N impurity: S = 1/2, gx,y = 2.0024, gz = 2.0024, Ax,y = 82 MHz, and Az = 114 MHz; surface impurities: S = 1/2, gx,y = 2.0029 and gz = 2.0027; and (ii) for the nitroxide component: S = 1/2, gx = 2.0086, gy = 2.0056, gz = 2.0033, Ax = 6.5, Ay = 5.6, Az = 37 MHz; and spectral linewidth = 0.5 mT. By analyzing the best-fit spectra obtained from simulations at both the X-band (9.3 GHz) and 115 GHz, the rotational correlation time of the ND-tethered TEMPO (τc) was determined to be 4.1 ns.
S11
Figure S17. EPR spectra of TEMPO-ND. Blue solid lines represent the experimental data, red dotted lines are the simulated spectra, green solid lines are ND components, and magenta solid lines represent the nitroxide components. (a) X-band (~9.3 GHz) spectrum. The nitroxide signal was magnified by 2 times to show the EPR lineshape. (b) 115 GHz EPR spectrum. The nitroxide signal was magnified by 10 times to show the EPR lineshape.
References 1. Grant, G. P. G. and Qin, P. Z., A facile method for attaching nitroxide spin labels at the 50 terminus of nucleic acids. Nucl. Acids Res. 2007, 35, 1-8. 2. Stoll, S. and Schweiger, A., EasySpin, a comprehensive software package for spectral simulation and analysis in EPR. J. Magn. Reson. 2006, 178, 42-55. 3. Romanova, E. E., Akiel, R., Cho, F. H. and Takahashi, S., Grafting nitroxide radicals on nanodiamond surface using click chemistry. J. Phys. Chem. A 2013, 117, 11933–11939. 4. Popova, A. M., Kálai, T., Hideg, K. and Qin, P. Z., Site-specific DNA structural and dynamic features revealed by nucleotide-independent nitroxide probes. Biochemistry 2009, 48, 85408550. 5. Ding, Y., Zhang, X., Tham, K. W. and Qin, P. Z., Experimental mapping of DNA duplex shape enabled by global lineshape analyses of a nucleotide-independent nitroxide probe. Nucl. Acids Res. 2014, 42, e140. 6. Cho, F. H., Stepanov, V., Abeywardana, C. and Takahashi, S., 230/115 GHz electron paramagnetic resonance/double electron-electron resonance spectroscopy. Methods Enzymol. 2015, 563, 95-118. 7. Cho, F. H., Stepanov, V. and Takahashi, S., A high-frequency electron paramagnetic resonance spectrometer for multi-dimensional, -frequency and -phase pulsed measurements. Rev. Sci. Instrum. 2014, 85, 075110.
S12
