- Email: [email protected]
A different approach to multiplicity-edited heteronuclear single quantum correlation spectroscopy
Accepted Manuscript A Different Approach to Multiplicity-Edited Heteronuclear Single Quantum Correlation Spectroscopy Peyman Sakhaii, Wolfgang Bermel PII: DOI: Reference:
S1090-7807(15)00158-5 http://dx.doi.org/10.1016/j.jmr.2015.07.006 YJMRE 5677
To appear in:
Journal of Magnetic Resonance
Received Date: Revised Date:
1 June 2015 21 July 2015
Please cite this article as: P. Sakhaii, W. Bermel, A Different Approach to Multiplicity-Edited Heteronuclear Single Quantum Correlation Spectroscopy, Journal of Magnetic Resonance (2015), doi: http://dx.doi.org/10.1016/j.jmr. 2015.07.006
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
A Different Approach to MultiplicityEdited Heteronuclear Single Quantum Correlation Spectroscopy
Peyman Sakhaii NMR Laboratory of SANOFI, C&BD (Chemistry & Biotechnology Development Frankfurt Chemistry) Industriepark Hoechst, Building G849, D-65926 Frankfurt/Main, Germany E-mail: [email protected] Tel: +49 (0) 69 305 21875
Wolfgang Bermel Bruker BioSpin GmbH, Silberstreifen, D-76287 Rheinstetten, Germany E-mail: [email protected] Tel: +49 (0) 721 5161 6119
Abstract
A new experiment for recording multiplicity-edited HSQC spectra is presented. In standard multiplicity-edited HSQC experiments, the amplitude of CH2 signals is negative compared to those of CH and CH3 groups. We propose to reverse the sign of 13C frequencies of CH2 groups in t1 as criteria for editing. Basically, a modified [BIRD]r,x element (Bilinear Rotation Pulses and Delays) is inserted in a standard HSQC pulse sequence with States-TPPI frequency detection in t1 for this purpose. The modified BIRD element was designed in such a way as to pass or stop the evolution of the heteronuclear 1JHC coupling. This is achieved by adding a 180° proton RF pulse in each of the 1/2J periods. Depending on their position the evolution is switched on or off. Usually, the BIRDelement is applied on real and imaginary increments of a HSQC experiment to achieve the editing between multiplicities. Here, we restrict the application of the modified BIRD element to either real or imaginary increments of the HSQC. With this new scheme for editing, changing the frequency and / or amplitude of the CH2 signals becomes available. Reversing the chemical shift axis for CH2 signals simplifies overcrowded frequency regions and thus avoids accidental signal cancellation in conventional edited HSQC experiments. The practical implementation is demonstrated on the protein Lysozyme. Advantages and limitations of the idea are discussed.
Keywords HSQC, multiplicity-edited HSQC, BIRD element, Inverse heteronuclear correlation, States-TPPI, IP / AP signal detection
Introduction
The importance of being able to edit the NMR signals of heteronuclei by the number of attached protons was discovered very early. Among the most popular editing experiments are those like DEPT, APT, SEFT and PENDANT based on multiple pulse techniques and heteronuclear detection [1 and references cited therein]. In cases where the sample amount is not the limiting factor, the DEPT 13C NMR experiment became routine in NMR laboratories. However, once sample amount is an issue, due to the low gyromagnetic ratio of 13C nuclei this is not an option. This motivated the development of editing filter in combination with proton detection. Because of the higher gyromagnetic ratio of protons, a better sensitivity is achieved. The resulting multiplicity-edited gradient aided HSQC experiment has widely been accepted as a useful tool for structure elucidation of small molecules and medium sized peptides [2--10]. Therefore the experiment became indispensable in many industrial NMR laboratories [6-7]. The method combines proton detected heteronuclear chemical shift correlation with information about the number of protons attached to the heteronuclei. In the 13C multiplicity-edited experiment, the editing is based on the difference in sign of the signals of CH, CH3 and CH2 spin systems. This is achieved by inserting an additional delay of 1/JHC. During this period the 2IzSy and 2IzSx coherence state of the heteronuclear spin will be modulated according to the number of attached protons [2-3]. One limitation of the 13C edited HSQC experiment is that for any given length of the additional delay the 1/JCH condition is only fulfilled for one particular value of J. Since the value of J varies quite significantly depending on the hybridisation of the carbon nucleus, this mismatch leads to a significant loss in signal intensity. So called matched adiabatic sweep pulse have been suggested to address the problem of inefficient refocusing. The approach uses adiabatic refocusing pulses that have sweep rates synchronized with the variation of 1JCH across the chemical shift range [8-11].
In multiplicity-edited HSQC experiments, the signal amplitude of a CH2 cross peaks appears negative and in case of strong signal overlap, there is a risk of signal cancellation at lower digital resolution. In this paper, we suggest a multiplicity-edited experiment addressing this problem. The proposed NMR experiment produces multiplicity-edited HSQC spectra, where the sign of the 13C frequencies of CH2 groups is reversed, compared to that of CH and CH3 groups. Depending on the actual implementation of the proposed scheme the sign of the signals from CH 2 groups can be inverted in addition.
Results and discussions
To achieve editing of proton multiplicities as described above, the pulse sequence depicted in figure 1 is utilized. Figure 1a shows a standard HSQC with a slightly modified [BIRD]r,x element [1214]. The sign of the heteronuclear frequency in the indirect dimension is determined by the States-TPPI protocol, leading to pure amplitude modulations of the real and imaginary increments [15]. Figure 1
After the first INEPT period, 2IzSy and 2IzSx coherences are created in real and imaginary increments respectively. An editing delay of 1/JHC is inserted at the end of t1 evolution period, whereby the effect of the editing function can be switched on and off, without changing the total length of the sequence (figure 1a-c). If the standard [BIRD]r,x element is modified by inserting a 180° proton RF pulse in the middle of each 1/2J period (figure 1b) the overall effect of the [BIRD]r,x element is turned off and no heteronuclear JCH coupling is evolving in the 1/J period. Alternatively, the position of the two 180° proton RF pulses can be shifted to the center of the [BIRD] r,x element
(figure 1c) ,again allowing the evolution of the heteronuclear JCH coupling and turning the editing function on (figure 1c). In summary, in the first case (figure 1b), the evolution of the heteronuclear J coupling is switched off resulting in a regular HSQC spectrum. In the second case (figure 1c) with editing being on, the HSQC signal is modulated by the number of the attached protons. Changing just the position of the 180° pulses has the advantage of creating an equal transverse relaxation (T2) contribution between the two operation modes. Now, we propose to execute the editing function on the real or imaginary increments, selectively. This is in contrast to conventional multiplicity-edited HSQC experiments, where the editing filter is executed on both real and imaginary increments. In the following, we restrict our discussion to the effects of applying the editing function to change the sign of the CH2 coherence state. In case of CH and CH3, the editing filter does not change the sign of their coherence state and therefore will not be discussed here. Table 1 summarizes possible combinations of applying the editing filter on real and imaginary increments to achieve the sign reversal of frequency and/or amplitude. Table 1 In principle, there are 4 possibilities to be discussed. In case A (table 1), there is no editing on real and imaginary increments. Therefore, the sign of 13C frequency and the signal amplitude remain unchanged. Case A corresponds to a simple HSQC spectrum with no editing. In case B, the editing function is executed on real and imaginary increments, producing a standard multiplicity-edited HSQC spectrum (figure 2). Figure 2
As expected, the amplitude of the CH2 signal is negative. In case C the editing filter is used exclusively to the imaginary increments. The real part of the increment was recorded with the editing switched off. As result of this combination the sign of 13C resonance frequency is reversed while keeping the sign of the amplitude unchanged (figure 3a). Finally, in case D, the editing filter
is applied only on the real increments, while the imaginary part of the increment was recorded without any editing (figure 3b). A multiplicity-edited spectrum is obtained, with the 13C frequencies and amplitudes for the CH2 groups reversed in sign (figure 3b). Figure 3
It is worth notingce that an imbalance in the editing element leads to ambiguous residual signals at the original frequencies. To avoid those, it is important to make sure that the real and imaginary signals exhibit comparable amplitudes. Our first trial to use simple 180° pulses applied on 1H and 13
C nuclei during the editing delay led to significant residual signal artefacts. Even the application
of [BIRD]d,x has shown slight imbalance in producing similar real and imaginary signal amplitudes. Best results could be obtained by a [BIRD]r,x element running in the two modes as depicted in figure 1a-c. Since the signals of CH3 groups are usually the most intense, artefacts will predominantly be visible in the methyl region. Hence the delay in the [BIRD]r,xelement was adjusted for a coupling constant of 125 Hz. With this, residual signals were detected with a maximum intensity of up to 4 % for few CH3 resonances. Given the signal to noise of the spectrum artefacts below about 2% would disappear in the noise. A quick check of the 1JHC coupling constants showed a variation of 124 to 130 Hz for the majority of the residues, but with values up to 138Hz in some cases. The appearance of large artefacts correlates with the large deviations from the 125 Hz for with the delays in the BIRD element were adjusted. Natural products may have structural elements which give rise to unusually large couplings constants. In this case artefacts associated with J mismatch will be an issue as well.
Experimental verification of the pulse sequence (Figure 1) was shown on a 4.4 mM solution of chicken egg lysozyme (129 amino acids, Figure 3a, 3b). The spectrum depicted in figure 3a demonstrates the proof of principle of editing on the imaginary increments (table 1; case C). As
expected, the sign of the 13C resonance frequencies of CH2 groups is reversed, compared to the sign of 13C frequencies of CH and CH3 (Figure 3a). The HSQC spectrum depicted in figure 3b is obtained by applying the editing filter on the real increments of the experiment (table 1; case D) inverting the amplitude of the cross peaks as well.
In case of strongly overlapping CH, CH3 and CH2 spectral regions, high digital resolution in the indirect dimension is required for the conventional multiplicity-edited HSQC experiment to avoid signal cancellation. The major advantage of the proposed experiment is that there is no need for very high digital resolution. By reversing the resonance position of CH2 groups, the cross peaks appear in an empty region of the HSQC spectrum. The empty region at the low field 13C and high field 1H part of the spectrum is usually free from any resonance. Thus signal cancellation is no longer an issue. This becomes even more important in those cases where signals cannot be separated at high digital resolution because of the similarity in chemical shift. Separating the signals into different regions of the spectrum will allow an unambiguous identification. Our experiences with the new sequence show, that strong signal overlaps between CH3, CH and CH2 are rapidly resolved after few increments in the indirect dimension, making the experiment very attractive for intrinsically t1 resolution limited samples. This limitation might be due to very short 13C relaxation times, where high acquisition times in t 1 are not accessible. Figure 4 To verify the capability of the proposed experiment, we reprocessed the standard edited HSQC (figure 4a) and its frequency reversed counterpart (figure 4b-c) at low digital resolution (64 complex data points in t1). The result is shown in Figure 4b-c. For example, the 1H and 13C resonances at 2.1 and 29 ppm are clearly visible in the frequency reversed HSQC (figure 4b). For the same spectral region, signal cancellation is observed in the standard edited HSQC (figure 4a). Also a number of signals around 1H and 13C frequencies of 1.6 and 40 ppm respectively are lost in
the standard edited HSQC experiment (figure 4a). In the experiment with frequency reversal those signals are clearly visible (figure 4b). Finally, the figure 4c is plotted to highlight the presence of a large number of CH2 resonances, which would otherwise be strongly obscured. The sensitivity of the multiplicity-edited frequency reversed edited HSQC experiment was determined by comparing its signal-to-noise ratio to that of a regular edited HSQC experiment. The values of CH, CH2 and CH3 signals from both experiments were determined and compared. The following ratios were found: CH (97 % ), CH aromatic (86 %), CH2 (88 %) and CH3 (86 %) the frequency reversed edited HSQC experiment thus in average shows 90% of the intensity of a regular edited HSQC.
Conclusion
We have introduced a new multiplicity-edited pulse sequence for obtaining edited HSQC spectra, where the resonance frequency and/or the amplitude of CH2 signal is reversed. For that, we have applied a modification to the [BIRD]r,x element to stop or pass the evolution of heteronuclear 1JHC coupling. This editing feature is applied to either real or imaginary increments of a HSQC experiment. The proposed experiment rapidly resolves CH2 cross peaks from overcrowded spectra without the need of high digital resolution along t1. Combining the present method with TOCSY mixing periods, opens up new possibilities for obtaining HSQC-TOCSY and other related hybrid experiments.
Figure caption 1 For obtaining multiplicity-edited frequency reversed HSQC spectra, the depicted pulse sequence is used. In sequence a) a regular HSQC pulse sequence with a modified [BIRD]r,x element is shown. Unless stated otherwise, pulses are applied along the x-axis. Narrow and wide filled rectangles represent non-selective 1H or 13C 90°and 180° pulses respectively. Open trapezoids represent smoothed chirp pulses [16] for inversion with a pulse length of 500 us, a sweep width of 60 kHz and 20% smoothing of the amplitude on either ends. The chirp pulse was defined with 1000 points and applied at a RF field strength (B1max/2) of 9.8 kHz (90° rectangular pulse of 25.5 us). This corresponds to a Q-factor of 5 (determined with ShapeTool). The pulsed field gradients are indicated as filled sine envelopes and are 1 ms in length with an amplitude according to a smoothed rectangle (10% smoothing on either side). They are used as crusher gradients for artefact suppression. The amplitudes of the gradient pulses have the following ratio: G 1 = 63%, G2 = 48%, G3 = 35 (with 100% being 53.5 G/cm). The pulsed field gradients are applied along the z-axis followed by a gradient recovery delay of 200 us. The following phase cycling was used for the pulse sequences: 3 = 0 2, 4 = 0 0 2 2, rec = 0 2 2 0. The multiplicity-edited HSQC spectrum was recorded using 4 scans per increment and 2 s relaxation delay. The following delay parameter was used: = 1 / (4 JCH) = 1.72 ms, with JCH = 145 Hz; = 1 / (2 JCH) = 4 ms, with JCH = 125 Hz; initial value for t1 increment was 3 us. Frequency discrimination in the 13C evolution dimension is achieved by States-TPPI protocol. Broadband adiabatic 13C decoupling [16] was applied during the acquisition (t2) using a 1.5 ms chirp pulse at a RF field strength (B1max/2) of 1.8 kHz with a p5m4 decoupling supercycle.
Table caption 1 Table 1 summarizes the editing possibilities of the modified [BIRD]r,x element on the 2IzSy
magnetization of CH2 groups. Real and imaginary increments in t1 are recorded according to States-TPPI protocol. A) No editing on real and imaginary increments; B) editing on real and imaginary increments; C) editing only on imaginary increments, leading to a reversed sign of 13C magnetization of CH2 groups; D) editing only on real increments, leading to a reversed sign of 13C frequency and the signal amplitude.
Figure caption 2 The figure shows the result of a regular multiplicity-edited HSQC spectrum of a 4.4 mM solution of commercially available chicken egg lysozyme (129 amino acid) in 20% acetic acid-d4/80% D2O. The experiment uses adiabatic refocusing pulses where the sweep rate is synchronized with the variation of 1JCH across the chemical shift range [10-11]. This compensates refocusing inefficiency, thus minimising loss of sensitivity. The spectrum was acquired for the purpose of comparison with the spectra obtained by the new pulse sequence in figure 1.
Figure caption 3 The spectra were collected on a 4.4 mM solution of the commercially available chicken egg Lysozyme (129 amino acid) in 20% acetic acid-d4/80% D2O utilizing the pulse sequence depicted in figure 1. The spectra demonstrate the result of applying the modified [BIRD]r,x element. Figure 3a) shows a spectrum with reversed sign of the 13C resonance frequencies of CH2 groups, while keeping the signal amplitude positive. The modified [BIRD]r,x element was applied exclusively on the imaginary increments of the experiment. In Figure 3b) the [BIRD]r,x editing function is applied on the real increments only. As a result, the sign of the 13C frequencies of CH2 groups and the signal amplitudes are reversed. All experiments were acquired on a Bruker-Avance II spectrometer (Bruker BioSpin, Rheinstetten, Germany) operating at 600.13 MHz proton frequency, equipped with a 5 mm triple resonance TCI-Cryoprobe and a z-axis pulsed field gradient accessory. All
spectra were processed with the processing software TopSpin 3.1 pl7. The temperature was kept at 300 K. No water suppression element was incorporated into the pulse sequence to suppress the residual water signal. The spectral width in direct 1H dimension was 12019.2 Hz leading to an acquisition time of 85.2 ms with a physical dwell time of 83.2 us (number of complex points: 1 k). The experiment used a 13C sweep width of 27.16 kHz (t2(max) of 18.8 ms with a FID resolution of 53.0 Hz/point, number of complex points: 512). Total measurement time of the reference and edited HSQC experiment was 2 h 30 min. In the observe dimension data were multiplied by a shifted sine squared function (shift of 90°) and zero filled. The 2D data set was forward linear predicted by a factor of two in the indirect dimension, apodised by multiplying with a shifted sine squared function (shift of 90°) and then zero-filled to yield a final matrix of 2048 (F2) x 1024 (F1) data points.
Figure caption 4 Figure 4 displays spectral expansions of a) the regular multiplicity-edited HSQC (Fig 2), b) the frequency reversed edited HSQC containing CH and CH3 cross peaks and c) frequency reversed edited HSQC containing the CH2 cross peaks (Fig 3a-b). For better comparison to the original resonance position of the CH2 groups, the spectrum shown in 4c (Fig 3b) was reversed in F1 by software.
References [1]
S. Braun, H.O. Kalinowski, S. Berger, 150 and more basic NMR experiments, A practical course 2nd expanded edition, Weinheim, New York, Chichester, Brisbane, Singapur, Toronto, Wiley-VCH (1998) 155-199.
[2]
L. E. Kay, A. Bax, Separation of NH and NH2 resonances in 1H-detected heteronuclear multiple-quantum correlation spectra, J. Magn. Reson. 84 (1989) 598-603.
[3]
D. G. Davis, Improved multiplet editing of proton-detected, heteronuclear shift-correlation spectra, J.Magn. Reson. 91 (1991) 665–672.
[4]
P. Schmieder, T. Domke, D. G. Norris, M. Kurz, H. Kessler, D. Leibfritz, Editing of multiplicity in two- and three dimensional heteronuclear NMR spectroscopy by fourier transformation of the pulse-angle dependency, J. Magn. Reson. 93 (1991) 430-435.
[5]
X. Zhang, C. Wang, 1H-detected editable heteronuclear multiple-quantum correlation experiment at natural abundance, J. Magn. Reson. 91 (1991) 618-623.
[6]
C. Griesinger, H. Schwalbe, J. Schleucher, M. Sattler, Proton-detected heteronuclear and multidimensional NMR, Chapter 3, Two Dimensional NMR spectroscopy, Application for chemists and biochemists, Second edition, W. R. Croasmun, R. M. K. Carlson, (1994) 457580.
[7]
W. Wilker, D. Leibfritz, R. Kerssebaum, W. Bermel, Gradient selection in inverse heteronuclear correlation spectroscopy, Magn. Reson. Chem. 31 (1993) 287-292.
[8]
J.M. Boehlen, G. Bodenhausen, Experimental aspects of chirp NMR spectroscopy, J. Magn. Reson. A 102 (1993) 293–301.
[9]
E. Kupce, R. Freeman, compensation for spin-spin coupling effects during adiabatic pulses, J. Magn. Reson. 127 (1997) 36-48.
[10]
C. Zwahlen, P. Legault, S. J. F. Vincent, J. Greenblatt, R. Konrat, L. E. Kay, Methods for measurement of intermolecular NOEs by multinuclear NMR spectroscopy: Application to a bacteriophage L N-Peptide/box RNA complex, J. Am. Chem. Soc. 119 (1997) 6711-6721.
[11]
R.D. Boyer, R. Johnson, K. J Krishnamurthy., Compensation of refocusing inefficiency with synchronized inversion sweep (CRISIS) in multiplicity-edited HSQC, J. Magn. Reson. 165 (2003) 253-259.
[12]
J.R. Garbow, D.P. Weitekamp, A. Pines, Bilinear rotation decoupling of homonuclear scalar interactions, Chem. Phys. Lett. 93 (1982) 504–509.
[13]
D. Uhrin, T. Liptaj, K. Koever, Modified BIRD pulses and design of heteronuclear pulse sequences, J. Magn. Reson. A 101 (1993) 41–46.
[14]
A. Bax, Broadband homonuclear decoupling in heteronuclear shift correlation NMR spectroscopy, J. Magn. Reson. 53 (1983) 517–520.
[15]
D. J. States, R. A. Haberkorn, D. J. Ruben, A two-dimensional nuclear overhauser experiment with pure absorption phase in four quadrants, J. Magn. Reson. 48 (1982) 286292.
[16]
R. Fu, G. Bodenhausen, Broadband decoupling in NMR with frequency-modulated chirp pulses, Chem. Phys. Lett. 245 (1995) 415–420.
#
A B
C
D
Application of BIRD filter on 13 C transverse magnetization <2IzSy> No editing Editing on real and imaginary increments Editing on imaginary increments Editing on real increments
Real
Imaginary
Sign of cross peak amplitude
Sign of cross peak frequency < c>
+ cos ( c x t1) - cos ( c x t1)
+ sin ( c x t1) - sin ( c x t1)
Positive Negative
Positive Positive
+ cos ( c x t1)
- sin ( c x t1)
Positive
Negative
- cos ( c x t1)
+ sin ( c x t1)
Negative
Negative
Graphical abstract
Highlights
A different multiplicity edited HSQC with States-TPPI frequency detection in t1 is proposed. The sign of the 13C frequency of the signals of CH2 groups and / or their amplitude is reversed. A modified [BIRD]r,x filter is applied on real or imaginary increments in t1. Signal cancellation, caused by editing at low digital resolution in crowded regions is avoided. The CH2 cross peaks appear in the dark region of HSQC spectra minimising signal overlap.
S1090-7807(15)00158-5 http://dx.doi.org/10.1016/j.jmr.2015.07.006 YJMRE 5677
To appear in:
Journal of Magnetic Resonance
Received Date: Revised Date:
1 June 2015 21 July 2015
Please cite this article as: P. Sakhaii, W. Bermel, A Different Approach to Multiplicity-Edited Heteronuclear Single Quantum Correlation Spectroscopy, Journal of Magnetic Resonance (2015), doi: http://dx.doi.org/10.1016/j.jmr. 2015.07.006
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
A Different Approach to MultiplicityEdited Heteronuclear Single Quantum Correlation Spectroscopy
Peyman Sakhaii NMR Laboratory of SANOFI, C&BD (Chemistry & Biotechnology Development Frankfurt Chemistry) Industriepark Hoechst, Building G849, D-65926 Frankfurt/Main, Germany E-mail: [email protected] Tel: +49 (0) 69 305 21875
Wolfgang Bermel Bruker BioSpin GmbH, Silberstreifen, D-76287 Rheinstetten, Germany E-mail: [email protected] Tel: +49 (0) 721 5161 6119
Abstract
A new experiment for recording multiplicity-edited HSQC spectra is presented. In standard multiplicity-edited HSQC experiments, the amplitude of CH2 signals is negative compared to those of CH and CH3 groups. We propose to reverse the sign of 13C frequencies of CH2 groups in t1 as criteria for editing. Basically, a modified [BIRD]r,x element (Bilinear Rotation Pulses and Delays) is inserted in a standard HSQC pulse sequence with States-TPPI frequency detection in t1 for this purpose. The modified BIRD element was designed in such a way as to pass or stop the evolution of the heteronuclear 1JHC coupling. This is achieved by adding a 180° proton RF pulse in each of the 1/2J periods. Depending on their position the evolution is switched on or off. Usually, the BIRDelement is applied on real and imaginary increments of a HSQC experiment to achieve the editing between multiplicities. Here, we restrict the application of the modified BIRD element to either real or imaginary increments of the HSQC. With this new scheme for editing, changing the frequency and / or amplitude of the CH2 signals becomes available. Reversing the chemical shift axis for CH2 signals simplifies overcrowded frequency regions and thus avoids accidental signal cancellation in conventional edited HSQC experiments. The practical implementation is demonstrated on the protein Lysozyme. Advantages and limitations of the idea are discussed.
Keywords HSQC, multiplicity-edited HSQC, BIRD element, Inverse heteronuclear correlation, States-TPPI, IP / AP signal detection
Introduction
The importance of being able to edit the NMR signals of heteronuclei by the number of attached protons was discovered very early. Among the most popular editing experiments are those like DEPT, APT, SEFT and PENDANT based on multiple pulse techniques and heteronuclear detection [1 and references cited therein]. In cases where the sample amount is not the limiting factor, the DEPT 13C NMR experiment became routine in NMR laboratories. However, once sample amount is an issue, due to the low gyromagnetic ratio of 13C nuclei this is not an option. This motivated the development of editing filter in combination with proton detection. Because of the higher gyromagnetic ratio of protons, a better sensitivity is achieved. The resulting multiplicity-edited gradient aided HSQC experiment has widely been accepted as a useful tool for structure elucidation of small molecules and medium sized peptides [2--10]. Therefore the experiment became indispensable in many industrial NMR laboratories [6-7]. The method combines proton detected heteronuclear chemical shift correlation with information about the number of protons attached to the heteronuclei. In the 13C multiplicity-edited experiment, the editing is based on the difference in sign of the signals of CH, CH3 and CH2 spin systems. This is achieved by inserting an additional delay of 1/JHC. During this period the 2IzSy and 2IzSx coherence state of the heteronuclear spin will be modulated according to the number of attached protons [2-3]. One limitation of the 13C edited HSQC experiment is that for any given length of the additional delay the 1/JCH condition is only fulfilled for one particular value of J. Since the value of J varies quite significantly depending on the hybridisation of the carbon nucleus, this mismatch leads to a significant loss in signal intensity. So called matched adiabatic sweep pulse have been suggested to address the problem of inefficient refocusing. The approach uses adiabatic refocusing pulses that have sweep rates synchronized with the variation of 1JCH across the chemical shift range [8-11].
In multiplicity-edited HSQC experiments, the signal amplitude of a CH2 cross peaks appears negative and in case of strong signal overlap, there is a risk of signal cancellation at lower digital resolution. In this paper, we suggest a multiplicity-edited experiment addressing this problem. The proposed NMR experiment produces multiplicity-edited HSQC spectra, where the sign of the 13C frequencies of CH2 groups is reversed, compared to that of CH and CH3 groups. Depending on the actual implementation of the proposed scheme the sign of the signals from CH 2 groups can be inverted in addition.
Results and discussions
To achieve editing of proton multiplicities as described above, the pulse sequence depicted in figure 1 is utilized. Figure 1a shows a standard HSQC with a slightly modified [BIRD]r,x element [1214]. The sign of the heteronuclear frequency in the indirect dimension is determined by the States-TPPI protocol, leading to pure amplitude modulations of the real and imaginary increments [15]. Figure 1
After the first INEPT period, 2IzSy and 2IzSx coherences are created in real and imaginary increments respectively. An editing delay of 1/JHC is inserted at the end of t1 evolution period, whereby the effect of the editing function can be switched on and off, without changing the total length of the sequence (figure 1a-c). If the standard [BIRD]r,x element is modified by inserting a 180° proton RF pulse in the middle of each 1/2J period (figure 1b) the overall effect of the [BIRD]r,x element is turned off and no heteronuclear JCH coupling is evolving in the 1/J period. Alternatively, the position of the two 180° proton RF pulses can be shifted to the center of the [BIRD] r,x element
(figure 1c) ,again allowing the evolution of the heteronuclear JCH coupling and turning the editing function on (figure 1c). In summary, in the first case (figure 1b), the evolution of the heteronuclear J coupling is switched off resulting in a regular HSQC spectrum. In the second case (figure 1c) with editing being on, the HSQC signal is modulated by the number of the attached protons. Changing just the position of the 180° pulses has the advantage of creating an equal transverse relaxation (T2) contribution between the two operation modes. Now, we propose to execute the editing function on the real or imaginary increments, selectively. This is in contrast to conventional multiplicity-edited HSQC experiments, where the editing filter is executed on both real and imaginary increments. In the following, we restrict our discussion to the effects of applying the editing function to change the sign of the CH2 coherence state. In case of CH and CH3, the editing filter does not change the sign of their coherence state and therefore will not be discussed here. Table 1 summarizes possible combinations of applying the editing filter on real and imaginary increments to achieve the sign reversal of frequency and/or amplitude. Table 1 In principle, there are 4 possibilities to be discussed. In case A (table 1), there is no editing on real and imaginary increments. Therefore, the sign of 13C frequency and the signal amplitude remain unchanged. Case A corresponds to a simple HSQC spectrum with no editing. In case B, the editing function is executed on real and imaginary increments, producing a standard multiplicity-edited HSQC spectrum (figure 2). Figure 2
As expected, the amplitude of the CH2 signal is negative. In case C the editing filter is used exclusively to the imaginary increments. The real part of the increment was recorded with the editing switched off. As result of this combination the sign of 13C resonance frequency is reversed while keeping the sign of the amplitude unchanged (figure 3a). Finally, in case D, the editing filter
is applied only on the real increments, while the imaginary part of the increment was recorded without any editing (figure 3b). A multiplicity-edited spectrum is obtained, with the 13C frequencies and amplitudes for the CH2 groups reversed in sign (figure 3b). Figure 3
It is worth notingce that an imbalance in the editing element leads to ambiguous residual signals at the original frequencies. To avoid those, it is important to make sure that the real and imaginary signals exhibit comparable amplitudes. Our first trial to use simple 180° pulses applied on 1H and 13
C nuclei during the editing delay led to significant residual signal artefacts. Even the application
of [BIRD]d,x has shown slight imbalance in producing similar real and imaginary signal amplitudes. Best results could be obtained by a [BIRD]r,x element running in the two modes as depicted in figure 1a-c. Since the signals of CH3 groups are usually the most intense, artefacts will predominantly be visible in the methyl region. Hence the delay in the [BIRD]r,xelement was adjusted for a coupling constant of 125 Hz. With this, residual signals were detected with a maximum intensity of up to 4 % for few CH3 resonances. Given the signal to noise of the spectrum artefacts below about 2% would disappear in the noise. A quick check of the 1JHC coupling constants showed a variation of 124 to 130 Hz for the majority of the residues, but with values up to 138Hz in some cases. The appearance of large artefacts correlates with the large deviations from the 125 Hz for with the delays in the BIRD element were adjusted. Natural products may have structural elements which give rise to unusually large couplings constants. In this case artefacts associated with J mismatch will be an issue as well.
Experimental verification of the pulse sequence (Figure 1) was shown on a 4.4 mM solution of chicken egg lysozyme (129 amino acids, Figure 3a, 3b). The spectrum depicted in figure 3a demonstrates the proof of principle of editing on the imaginary increments (table 1; case C). As
expected, the sign of the 13C resonance frequencies of CH2 groups is reversed, compared to the sign of 13C frequencies of CH and CH3 (Figure 3a). The HSQC spectrum depicted in figure 3b is obtained by applying the editing filter on the real increments of the experiment (table 1; case D) inverting the amplitude of the cross peaks as well.
In case of strongly overlapping CH, CH3 and CH2 spectral regions, high digital resolution in the indirect dimension is required for the conventional multiplicity-edited HSQC experiment to avoid signal cancellation. The major advantage of the proposed experiment is that there is no need for very high digital resolution. By reversing the resonance position of CH2 groups, the cross peaks appear in an empty region of the HSQC spectrum. The empty region at the low field 13C and high field 1H part of the spectrum is usually free from any resonance. Thus signal cancellation is no longer an issue. This becomes even more important in those cases where signals cannot be separated at high digital resolution because of the similarity in chemical shift. Separating the signals into different regions of the spectrum will allow an unambiguous identification. Our experiences with the new sequence show, that strong signal overlaps between CH3, CH and CH2 are rapidly resolved after few increments in the indirect dimension, making the experiment very attractive for intrinsically t1 resolution limited samples. This limitation might be due to very short 13C relaxation times, where high acquisition times in t 1 are not accessible. Figure 4 To verify the capability of the proposed experiment, we reprocessed the standard edited HSQC (figure 4a) and its frequency reversed counterpart (figure 4b-c) at low digital resolution (64 complex data points in t1). The result is shown in Figure 4b-c. For example, the 1H and 13C resonances at 2.1 and 29 ppm are clearly visible in the frequency reversed HSQC (figure 4b). For the same spectral region, signal cancellation is observed in the standard edited HSQC (figure 4a). Also a number of signals around 1H and 13C frequencies of 1.6 and 40 ppm respectively are lost in
the standard edited HSQC experiment (figure 4a). In the experiment with frequency reversal those signals are clearly visible (figure 4b). Finally, the figure 4c is plotted to highlight the presence of a large number of CH2 resonances, which would otherwise be strongly obscured. The sensitivity of the multiplicity-edited frequency reversed edited HSQC experiment was determined by comparing its signal-to-noise ratio to that of a regular edited HSQC experiment. The values of CH, CH2 and CH3 signals from both experiments were determined and compared. The following ratios were found: CH (97 % ), CH aromatic (86 %), CH2 (88 %) and CH3 (86 %) the frequency reversed edited HSQC experiment thus in average shows 90% of the intensity of a regular edited HSQC.
Conclusion
We have introduced a new multiplicity-edited pulse sequence for obtaining edited HSQC spectra, where the resonance frequency and/or the amplitude of CH2 signal is reversed. For that, we have applied a modification to the [BIRD]r,x element to stop or pass the evolution of heteronuclear 1JHC coupling. This editing feature is applied to either real or imaginary increments of a HSQC experiment. The proposed experiment rapidly resolves CH2 cross peaks from overcrowded spectra without the need of high digital resolution along t1. Combining the present method with TOCSY mixing periods, opens up new possibilities for obtaining HSQC-TOCSY and other related hybrid experiments.
Figure caption 1 For obtaining multiplicity-edited frequency reversed HSQC spectra, the depicted pulse sequence is used. In sequence a) a regular HSQC pulse sequence with a modified [BIRD]r,x element is shown. Unless stated otherwise, pulses are applied along the x-axis. Narrow and wide filled rectangles represent non-selective 1H or 13C 90°and 180° pulses respectively. Open trapezoids represent smoothed chirp pulses [16] for inversion with a pulse length of 500 us, a sweep width of 60 kHz and 20% smoothing of the amplitude on either ends. The chirp pulse was defined with 1000 points and applied at a RF field strength (B1max/2) of 9.8 kHz (90° rectangular pulse of 25.5 us). This corresponds to a Q-factor of 5 (determined with ShapeTool). The pulsed field gradients are indicated as filled sine envelopes and are 1 ms in length with an amplitude according to a smoothed rectangle (10% smoothing on either side). They are used as crusher gradients for artefact suppression. The amplitudes of the gradient pulses have the following ratio: G 1 = 63%, G2 = 48%, G3 = 35 (with 100% being 53.5 G/cm). The pulsed field gradients are applied along the z-axis followed by a gradient recovery delay of 200 us. The following phase cycling was used for the pulse sequences: 3 = 0 2, 4 = 0 0 2 2, rec = 0 2 2 0. The multiplicity-edited HSQC spectrum was recorded using 4 scans per increment and 2 s relaxation delay. The following delay parameter was used: = 1 / (4 JCH) = 1.72 ms, with JCH = 145 Hz; = 1 / (2 JCH) = 4 ms, with JCH = 125 Hz; initial value for t1 increment was 3 us. Frequency discrimination in the 13C evolution dimension is achieved by States-TPPI protocol. Broadband adiabatic 13C decoupling [16] was applied during the acquisition (t2) using a 1.5 ms chirp pulse at a RF field strength (B1max/2) of 1.8 kHz with a p5m4 decoupling supercycle.
Table caption 1 Table 1 summarizes the editing possibilities of the modified [BIRD]r,x element on the 2IzSy
magnetization of CH2 groups. Real and imaginary increments in t1 are recorded according to States-TPPI protocol. A) No editing on real and imaginary increments; B) editing on real and imaginary increments; C) editing only on imaginary increments, leading to a reversed sign of 13C magnetization of CH2 groups; D) editing only on real increments, leading to a reversed sign of 13C frequency and the signal amplitude.
Figure caption 2 The figure shows the result of a regular multiplicity-edited HSQC spectrum of a 4.4 mM solution of commercially available chicken egg lysozyme (129 amino acid) in 20% acetic acid-d4/80% D2O. The experiment uses adiabatic refocusing pulses where the sweep rate is synchronized with the variation of 1JCH across the chemical shift range [10-11]. This compensates refocusing inefficiency, thus minimising loss of sensitivity. The spectrum was acquired for the purpose of comparison with the spectra obtained by the new pulse sequence in figure 1.
Figure caption 3 The spectra were collected on a 4.4 mM solution of the commercially available chicken egg Lysozyme (129 amino acid) in 20% acetic acid-d4/80% D2O utilizing the pulse sequence depicted in figure 1. The spectra demonstrate the result of applying the modified [BIRD]r,x element. Figure 3a) shows a spectrum with reversed sign of the 13C resonance frequencies of CH2 groups, while keeping the signal amplitude positive. The modified [BIRD]r,x element was applied exclusively on the imaginary increments of the experiment. In Figure 3b) the [BIRD]r,x editing function is applied on the real increments only. As a result, the sign of the 13C frequencies of CH2 groups and the signal amplitudes are reversed. All experiments were acquired on a Bruker-Avance II spectrometer (Bruker BioSpin, Rheinstetten, Germany) operating at 600.13 MHz proton frequency, equipped with a 5 mm triple resonance TCI-Cryoprobe and a z-axis pulsed field gradient accessory. All
spectra were processed with the processing software TopSpin 3.1 pl7. The temperature was kept at 300 K. No water suppression element was incorporated into the pulse sequence to suppress the residual water signal. The spectral width in direct 1H dimension was 12019.2 Hz leading to an acquisition time of 85.2 ms with a physical dwell time of 83.2 us (number of complex points: 1 k). The experiment used a 13C sweep width of 27.16 kHz (t2(max) of 18.8 ms with a FID resolution of 53.0 Hz/point, number of complex points: 512). Total measurement time of the reference and edited HSQC experiment was 2 h 30 min. In the observe dimension data were multiplied by a shifted sine squared function (shift of 90°) and zero filled. The 2D data set was forward linear predicted by a factor of two in the indirect dimension, apodised by multiplying with a shifted sine squared function (shift of 90°) and then zero-filled to yield a final matrix of 2048 (F2) x 1024 (F1) data points.
Figure caption 4 Figure 4 displays spectral expansions of a) the regular multiplicity-edited HSQC (Fig 2), b) the frequency reversed edited HSQC containing CH and CH3 cross peaks and c) frequency reversed edited HSQC containing the CH2 cross peaks (Fig 3a-b). For better comparison to the original resonance position of the CH2 groups, the spectrum shown in 4c (Fig 3b) was reversed in F1 by software.
References [1]
S. Braun, H.O. Kalinowski, S. Berger, 150 and more basic NMR experiments, A practical course 2nd expanded edition, Weinheim, New York, Chichester, Brisbane, Singapur, Toronto, Wiley-VCH (1998) 155-199.
[2]
L. E. Kay, A. Bax, Separation of NH and NH2 resonances in 1H-detected heteronuclear multiple-quantum correlation spectra, J. Magn. Reson. 84 (1989) 598-603.
[3]
D. G. Davis, Improved multiplet editing of proton-detected, heteronuclear shift-correlation spectra, J.Magn. Reson. 91 (1991) 665–672.
[4]
P. Schmieder, T. Domke, D. G. Norris, M. Kurz, H. Kessler, D. Leibfritz, Editing of multiplicity in two- and three dimensional heteronuclear NMR spectroscopy by fourier transformation of the pulse-angle dependency, J. Magn. Reson. 93 (1991) 430-435.
[5]
X. Zhang, C. Wang, 1H-detected editable heteronuclear multiple-quantum correlation experiment at natural abundance, J. Magn. Reson. 91 (1991) 618-623.
[6]
C. Griesinger, H. Schwalbe, J. Schleucher, M. Sattler, Proton-detected heteronuclear and multidimensional NMR, Chapter 3, Two Dimensional NMR spectroscopy, Application for chemists and biochemists, Second edition, W. R. Croasmun, R. M. K. Carlson, (1994) 457580.
[7]
W. Wilker, D. Leibfritz, R. Kerssebaum, W. Bermel, Gradient selection in inverse heteronuclear correlation spectroscopy, Magn. Reson. Chem. 31 (1993) 287-292.
[8]
J.M. Boehlen, G. Bodenhausen, Experimental aspects of chirp NMR spectroscopy, J. Magn. Reson. A 102 (1993) 293–301.
[9]
E. Kupce, R. Freeman, compensation for spin-spin coupling effects during adiabatic pulses, J. Magn. Reson. 127 (1997) 36-48.
[10]
C. Zwahlen, P. Legault, S. J. F. Vincent, J. Greenblatt, R. Konrat, L. E. Kay, Methods for measurement of intermolecular NOEs by multinuclear NMR spectroscopy: Application to a bacteriophage L N-Peptide/box RNA complex, J. Am. Chem. Soc. 119 (1997) 6711-6721.
[11]
R.D. Boyer, R. Johnson, K. J Krishnamurthy., Compensation of refocusing inefficiency with synchronized inversion sweep (CRISIS) in multiplicity-edited HSQC, J. Magn. Reson. 165 (2003) 253-259.
[12]
J.R. Garbow, D.P. Weitekamp, A. Pines, Bilinear rotation decoupling of homonuclear scalar interactions, Chem. Phys. Lett. 93 (1982) 504–509.
[13]
D. Uhrin, T. Liptaj, K. Koever, Modified BIRD pulses and design of heteronuclear pulse sequences, J. Magn. Reson. A 101 (1993) 41–46.
[14]
A. Bax, Broadband homonuclear decoupling in heteronuclear shift correlation NMR spectroscopy, J. Magn. Reson. 53 (1983) 517–520.
[15]
D. J. States, R. A. Haberkorn, D. J. Ruben, A two-dimensional nuclear overhauser experiment with pure absorption phase in four quadrants, J. Magn. Reson. 48 (1982) 286292.
[16]
R. Fu, G. Bodenhausen, Broadband decoupling in NMR with frequency-modulated chirp pulses, Chem. Phys. Lett. 245 (1995) 415–420.
#
A B
C
D
Application of BIRD filter on 13 C transverse magnetization <2IzSy> No editing Editing on real and imaginary increments Editing on imaginary increments Editing on real increments
Real
Imaginary
Sign of cross peak amplitude
Sign of cross peak frequency < c>
+ cos ( c x t1) - cos ( c x t1)
+ sin ( c x t1) - sin ( c x t1)
Positive Negative
Positive Positive
+ cos ( c x t1)
- sin ( c x t1)
Positive
Negative
- cos ( c x t1)
+ sin ( c x t1)
Negative
Negative
Graphical abstract
Highlights
A different multiplicity edited HSQC with States-TPPI frequency detection in t1 is proposed. The sign of the 13C frequency of the signals of CH2 groups and / or their amplitude is reversed. A modified [BIRD]r,x filter is applied on real or imaginary increments in t1. Signal cancellation, caused by editing at low digital resolution in crowded regions is avoided. The CH2 cross peaks appear in the dark region of HSQC spectra minimising signal overlap.