Elimination of Zero-Quantum artifacts and sensitivity enhancement in perfect echo based 2D NOESY

Highlights

• Modified perfect echo scheme is implemented during t₁ period of 2D NOESY.

• Creation of Zero Quantum artifact is prevented for shorter t₁ period.

• For longer t₁ period a partial decoupling effect exists.

• Perfect echo along with a Zero-Quantum filter is applied.

• Many PE-NOE cross peaks displayed partial sensitivity enhancement in diluted spin systems.

Abstract

Zero-Quantum artifacts seriously degrade the performance of 2D NOESY. Homonuclear J-evolution during t₁ period generates Zero-Quantum and other higher quantum coherences which represent the magnetization loss and the artifacts created. We demonstrate that creation of such artifacts itself can be prevented for shorter t₁ period by a perfect echo based decoupling technique during t₁ period in a single scan. This is in contrast to existing methods that create unwanted coherence, and subsequently suppress that to produce a clean spectrum with a sensitivity penalty. Although decoupling performance of the present scheme remains robust for echo time 2τ short compared to 1/2J, we show that even a partial decoupling effect for extended t₁ (=2τ) period up to 100 ms along with a Zero-Quantum filter generates NOE spectrum from Cyclosporine A, in which majority of the cross peaks displayed partial sensitivity enhancement with few exceptions. However, in crowded proton spin systems like menthol, the enhancements were not observed and perfect echo NOESY displays similar performance as Zero-Quantum filtered NOESY.

Introduction

Homonuclear J-modulation produces J-multiplets which lowers resolution and sensitivity of ¹H signals. Homonuclear scalar decoupling can replace these multiplets with singlets offering high resolution, enhanced sensitivity, and suppression of certain artifacts. J-modulation can be suppressed by resonance-specific selective irradiation [1], rapid refocusing in the Carr–Purcell–Meiboom–Gill (CPMG) experiment [2], and slow refocusing in conjunction with cumulative pulse imperfections [3], [4], [5], [6], [7]. Other approaches are “perfect echo” (PE) approach for a weakly coupled AX spin system [8], [9] and their recent applications to arbitrary spin systems [10], [11], [12], [13]; homonuclear decoupling by supercycled BIRD (Bilinear rotation decoupling) blocks [14], [15], [16]; constant time evolution to remove the effect of J-coupling from the indirect t₁ dimension [17], [18], [19]; use of spatially-selective ¹H spin inversions in Zangger/Sterk sequence [20], [21], [22] and their improved versions for 2D NMR [23], [24]; and ultrahigh resolution method [25].

An undesired consequence of homonuclear J-evolution during t₁ period of 2D NOESY is the appearance of Zero-Quantum (ZQ) artifacts. The NOESY experiment is a versatile tool for structure elucidation of molecules in solution. The intensity of NOE cross peaks provides through space distance information between protons and arises from z-magnetization transfer via dipole–dipole cross relaxation during the mixing period [26], [27]. Ideally, only the z-magnetization should contribute to the spectrum, and either phase cycling or gradient is employed to ensure that. However, artifacts that appear via ZQ pathway cannot be discriminated from z-magnetization by such routine procedures.

The ZQ peaks have antiphase dispersive lineshape in both dimensions. Since NOE and ZQ peaks have identical frequencies in F₁ and F₂ dimensions, accurate integration of cross peaks are hampered by such artifacts. The long dispersive tails prevent accurate integration of other cross peaks, reduce effective resolution and introduce unwanted correlations. A number of methods have been proposed [23], [28], [29], [30], [31], [32], [33], [34] to suppress ZQ coherence. Most of these techniques focus on suppressing the ZQ coherence once it is already there, and therefore, pay sensitivity penalty. There exist very few approaches that attempt to prevent creation of such coherences. One such noteworthy approach utilizes band selective homonuclear decoupling in both dimensions of 2D NOESY for enhancing sensitivity and suppressing ZQ coherence [30]. Another such approach has been demonstrated for pure shift TOCSY utilizing Zangger/Sterk element which has inherent sensitivity penalty of 1–2 orders of magnitude [23]. Broadband ${\omega }_1$ decoupling in 2D NOE has also been carried out by time reversal of the evolution under scalar spin–spin interactions [29]. However, this experiment pays sensitivity penalty by a factor of 2.5 in comparison to conventional NOESY. A universal broadband homodecoupling in t₁ dimension of homonuclear correlation experiment has not been realized till date without a sensitivity penalty.

In conventional NOESY, homonuclear J-modulation during t₁ creates antiphase coherences which lead to unwanted ZQ and other higher order coherences. A broadband homo-decoupled t₁ period can prevent creation of antiphase coherences. This will raise the sensitivity of NOE cross peaks for protons involved in J-coupling. In the present work, we demonstrate a modification of the original perfect echo sequence [8], [9] that allows chemical shift encoding as well as homo-decoupling during t₁ period of 2D NOESY and eliminates the creation of unwanted coherences for shorter t₁ period. A similar approach was utilized for removal of J-cross peaks in 2D CAMELSPIN experiment in 1988 [35] before the original report of the perfect echo sequence. The perfect echo based J-refocusing works perfectly for a two spin system and partially for multiply coupled spins [9]. All the amide protons in peptides and proteins experience only one scalar coupling to ¹H^α except in glycine residues. For ¹H^α protons, apart from the amide protons, often there are either one or two protons at C^β except in alanine residues. Therefore, these systems do not represent very complicated spin system. Further, all these amide and H-alpha protons experience small vicinal coupling rather than large geminal couplings. A gain in in-phase intensity even in multiply coupled spin systems is reported for long echo time close to 3/4 J [9] using perfect echo sequence. Considering these developments, herein, we access the performance of ZQ suppression by a modified perfect echo based shorter t₁ period. Later, we explore the sensitivity enhancement of NOE cross peaks by a partial J-refocusing effect of the perfect echo combined with a ZQ filter for longer t₁ period of 50 and 100 ms