Although organic chemists typically focus on hydrogen and carbon when
analyzing complex molecules, there is an increasing amount of analysis
done on other elements such as phosphorous or sulfur. NMR is an
underappreciated tool to understand the role of these heteroatoms,
especially phosphorous, whose NMR-active isotopes are abundant and have
high gyromagnetic ratios.
Adenosine triphosphate (ATP) is ubiquitous in eukaryotic and prokaryotic cells and serves as an energy source. Its three phosphate groups and purine ring serve as a great template for some interesting NMR experiments and demonstration of how heteronuclear and heteronuclear 2-dimensional experiments can be used to elucidate molecular structure.
Figure 1: 1H NMR spectrum of 10 mg/ml ATP in D2O.
The 1H NMR spectrum of ATP is straightforward to acquire as long as one doesn’t mind losing the exchangable protons to the solvent. Water suppression techniques are unnecessary given the relatively sharp solvent signal (more on water suppression in a future post). The small peaks at δ 3.5 and d 1.2 ppm are residual triethylamine from the commercial material. Assignment of the non-aromatic peaks (H-1′, H-2′, H-3′, H-4′, H-5′) can be achieved using COSY but will not be shown here – see previous post on clarithromycin for details on 1H assignment using COSY.
Figure 2: 13C NMR spectrum, with the doublet at δ 84 ppm shown in inset
Likewise, the 13C NMR of ATP is straightforward to acquire with an overnight scan. A small amount of line broadening (LB = 1.0 Hz) was applied to improve the signal-to-noise, but limited to keep the maintain resolution of the doublet at δ 84 ppm (which arises due to 13C-31P coupling).
Figure 3: 1H-13C HSQC spectrum
The alkyl 13C resonances can easily be assigned using an HSQC spectrum, which correlates the 1H spectrum with the 13C spectrum. In this phase-sensitive version of the HSQC experiment, red cross-peaks have CH2 multiplicity (H-5′ / C-5′) while black cross-peaks have CH or CH3 multiplicity. The 13C resonances at δ 142 and d 145 ppm are C-2 and C-8 (but assignment not yet known), while the resonances at δ 118, 148, and 150 ppm do not have protons bonded (C-4, C-5, and C-6; assignment not yet known).
Figure 4: 1H-13C HMBC spectrum
Heteronuclear Multiple Bond Correlation is similar to the HSQC, except that 1-bond correlations are canceled out, and only 2- and 3-bond correlations are shown. The HMBC experiment allows one to identify all of the aromatic resonances, on both the 1H and 13C spectra. The H-1′ (δ 6.0 ppm) signal show correlations to the C-4 (δ 148 ppm) and C-8 (δ 142 ppm) signals. The H-8 (δ 8.5 ppm) signal shows correlations to the C-4 and C-5 (δ 118 ppm) signals. The H-2 (δ 8.3 ppm) signal shows correlations to the C-4 and C-6 (δ 150 ppm) signals, as well as a weak signal to C-5.
Now that the 1H and 13C spectra are fully assigned, we can direct our attention to the 31P NMR spectra.
Figure 5: 31P NMR spectrum
31P has decent sensitivity and high abundance and is a good nucleus to use as an example of heteronuclear spectroscopy. A 1-dimensional 31P spectrum shows three signals at δ -10, -11, and -23 ppm. This spectrum is 1H-coupled, but the 3-bond 1H-31P coupling is too small to see. The observed couplings are 31P-31P couplings, and based on the multiplicity, it is clear that the middle phosphate gives rise to the signal at δ -23 ppm. The other two signals can be differentiated using HMBC, except this time 1H-31P instead of 1H-13C.
As an internal standard, approximately 0.5 equivalents of trimethylphosphonoacetate were added. The 31P signal at δ +26.06 ppm gives strong crosspeaks to the methylene (δ 3.2 ppm) as well as methyl (δ 3.8 ppm) protons and helps calibrate the spectrum along the 31P axis.
Figure 6: 1H-31P HMBC spectrum, full (top) and enlarged on the ATP signals (bottom).
In this case, crosspeaks between the 31P signal at δ -11.27 ppm and the H-4′ and H-5′ signals from δ 4.15-4.35 ppm were observed. The crosspeaks establish the 31P signal at δ -11.27 ppm as the first phosphate, allowing assignment of the 31P NMR spectrum as well as confirmation that the triphosphate group is covalently attached to the nucleoside via that 5′-OH.
Adenosine triphosphate (ATP) is ubiquitous in eukaryotic and prokaryotic cells and serves as an energy source. Its three phosphate groups and purine ring serve as a great template for some interesting NMR experiments and demonstration of how heteronuclear and heteronuclear 2-dimensional experiments can be used to elucidate molecular structure.
Figure 1: 1H NMR spectrum of 10 mg/ml ATP in D2O.
The 1H NMR spectrum of ATP is straightforward to acquire as long as one doesn’t mind losing the exchangable protons to the solvent. Water suppression techniques are unnecessary given the relatively sharp solvent signal (more on water suppression in a future post). The small peaks at δ 3.5 and d 1.2 ppm are residual triethylamine from the commercial material. Assignment of the non-aromatic peaks (H-1′, H-2′, H-3′, H-4′, H-5′) can be achieved using COSY but will not be shown here – see previous post on clarithromycin for details on 1H assignment using COSY.
Figure 2: 13C NMR spectrum, with the doublet at δ 84 ppm shown in inset
Likewise, the 13C NMR of ATP is straightforward to acquire with an overnight scan. A small amount of line broadening (LB = 1.0 Hz) was applied to improve the signal-to-noise, but limited to keep the maintain resolution of the doublet at δ 84 ppm (which arises due to 13C-31P coupling).
Figure 3: 1H-13C HSQC spectrum
The alkyl 13C resonances can easily be assigned using an HSQC spectrum, which correlates the 1H spectrum with the 13C spectrum. In this phase-sensitive version of the HSQC experiment, red cross-peaks have CH2 multiplicity (H-5′ / C-5′) while black cross-peaks have CH or CH3 multiplicity. The 13C resonances at δ 142 and d 145 ppm are C-2 and C-8 (but assignment not yet known), while the resonances at δ 118, 148, and 150 ppm do not have protons bonded (C-4, C-5, and C-6; assignment not yet known).
Figure 4: 1H-13C HMBC spectrum
Heteronuclear Multiple Bond Correlation is similar to the HSQC, except that 1-bond correlations are canceled out, and only 2- and 3-bond correlations are shown. The HMBC experiment allows one to identify all of the aromatic resonances, on both the 1H and 13C spectra. The H-1′ (δ 6.0 ppm) signal show correlations to the C-4 (δ 148 ppm) and C-8 (δ 142 ppm) signals. The H-8 (δ 8.5 ppm) signal shows correlations to the C-4 and C-5 (δ 118 ppm) signals. The H-2 (δ 8.3 ppm) signal shows correlations to the C-4 and C-6 (δ 150 ppm) signals, as well as a weak signal to C-5.
Now that the 1H and 13C spectra are fully assigned, we can direct our attention to the 31P NMR spectra.
Figure 5: 31P NMR spectrum
31P has decent sensitivity and high abundance and is a good nucleus to use as an example of heteronuclear spectroscopy. A 1-dimensional 31P spectrum shows three signals at δ -10, -11, and -23 ppm. This spectrum is 1H-coupled, but the 3-bond 1H-31P coupling is too small to see. The observed couplings are 31P-31P couplings, and based on the multiplicity, it is clear that the middle phosphate gives rise to the signal at δ -23 ppm. The other two signals can be differentiated using HMBC, except this time 1H-31P instead of 1H-13C.
As an internal standard, approximately 0.5 equivalents of trimethylphosphonoacetate were added. The 31P signal at δ +26.06 ppm gives strong crosspeaks to the methylene (δ 3.2 ppm) as well as methyl (δ 3.8 ppm) protons and helps calibrate the spectrum along the 31P axis.
Figure 6: 1H-31P HMBC spectrum, full (top) and enlarged on the ATP signals (bottom).
In this case, crosspeaks between the 31P signal at δ -11.27 ppm and the H-4′ and H-5′ signals from δ 4.15-4.35 ppm were observed. The crosspeaks establish the 31P signal at δ -11.27 ppm as the first phosphate, allowing assignment of the 31P NMR spectrum as well as confirmation that the triphosphate group is covalently attached to the nucleoside via that 5′-OH.
No comments:
Post a Comment