Which COSY approach

Before moving on, we briefly examine the key characteristics of the different experiments and consider why these might be of interest to the research chemist. Table 5.5 summarises the most significant attributes, some of which have already been introduced 

Absolute-value (magnitude-mode) COSY-90 Sinqtle and robust, magnitude processing well suited to automated operation Phase-twisted Uneshapes produce poor resolution, which require strong resolution enhancement functions. Crosspeak fine structure not usually apparent 

Phase-sensitive COSY-90 High-resolution display due to absorptive Uneshapes. Crosspeak fine structure apparent J measurement possible Diagonal peaks have dispersive Uneshapes that may interfere with neighbouring crosspeaks. Requires high digital resolution to reveal multiple structures 

COSY-/3 Simple and robust. Magnimde processing weU suited to automated operation. SimpUfication of crosspeak structures reduces peak overlap. Vicinal and geminal coupUngs can be distinguished in some cases from tilt of peaks Usually requires magnitude-mode presentation as phase-sensitive variant has mixed-phase Uneshapes 

Before moving on we briefly examine the key characteristics of the different experiments and consider why these might be of interest to the research chemist. Table 5.5 summarises the most significant attributes, some of which have already been introduced whilst others are expanded in the sections that follow. Whilst the TOCSY experiment is not strictly a member of the COSY family, its information content is so closely related to that of COSY it has also been included in the table. 

---

Refinements of the technique can be used which eliminate all resonances from one- and two-spin systems, i.e., singlets and doublets (triple-quantum filtered COSY), and so on for higher-spin systems. The DQF-COSY experiment is not as sensitive as the normal COSY approach and the higher quantum filtering is usually achieved using pulse phase cycling with increments in phase angles of less than 90°, a facility not always available on older spectrometers. 

The order in which various NMR data are acquired is largely one of user preference. Acquisition of the proton reference spectrum will invariably be undertaken first. Whether a user next seeks to establish homo- or heteronuclear shift correlations is where individual preferences come into play. Many spectro-scopists proceed from the proton reference spectrum to either a COSY or a TOCS Y spectrum next, while others may prefer to establish direct proton-carbon chemical shift correlations. This author s preference is for the latter approach. From a multiplicity-edited HSQC spectrum you obtain not only the carbon chemical shifts, which give an indication of the location of heteroatoms, the degree of unsaturation and the like, but also the number of directly attached protons, which eliminates the need for the acquisition of a DEPT spectrum [51, 52]. The statement in the prior sentence presupposes, of course, that there the sensitivity losses associated with the acquisition of multiplicity-edited HSQC data are tolerable. 

The structures of the compounds were elucidated by a combination of NMR techniques (lH-, 13C-, and 13C-DEPT NMR) and chemical transformation, enzymatic degradation, and as well as mass spectrometry, which gives information on the saccharide sequence. A more recent approach consists of an extensive use of high-resolution 2D NMR techniques, such as homonuclear and heteronuclear correlated spectroscopy (DQF-COSY, HOHAHA, HSQC, HMBC) and NOE spectroscopy (NOESY, ROESY), which now play the most important role in the structural elucidation of intact glycosides. These techniques are very sensitive and non destructive and allow easy recovery of the intact compounds for subsequent biological testing. 

Broadband Hartmann-Hahn transfer can also be of assistance in alternative approaches to determine coupling constants that do not rely on E.COSY-type multiplets that are separated by large one-bond couplings. The homonuclear two-dimensional PICSY (pure in-phase correlation spectroscopy) experiment (Vincent et al., 1992, 1993), which is based on selective Hartmann-Hahn transfer using doubly selective irradiation, can... 

Intensity-based methods. 1H— H RDCs have been obtained by analyzing the intensity ratios of the diagonal and cross-peaks in a series of 2D CT COSY spectra.177 This method can only be applied to resolved resonances, for example, those of anomeric protons.149 Similar limitations apply to /-modulated ID directed COSY,153,178 which uses selective 180° pulses to produce a series of ID spectra for each pair of coupled spins. This approach has recently been extended to include additional selection blocks yielding a versatile method for the measurement of coupling constants in compounds with severely overlapping proton resonances such as those found in carbohydrates.154 The problem of overlapping resonances can also be resolved by involving 13C nuclei, as demonstrated on natural abundance (13C COSMO HSQC)32 or uniformly 13C isotopically enriched carbohydrates (2D-HSQC-(sel C, sel H)-CT COSY experiment).73,158... 

There are two general approaches through which stage (1), resonance assignment, can be accomplished, and the choice between them is determined essentially by the molecular mass of the protein. The first approach, which was pioneered in the laboratory of Kurt Wiithrich34 and which led to his award of the 2002 Nobel Prize in Chemistry, involves the use of 2D 1H—1 H (homonuclear) NMR experiments such as COSY, NOESY, and TOCSY (see below). This approach is still widely used today, but only for proteins smaller than lOkDa that cannot be isotopically labeled. Nevertheless, the homonuclear NMR strategy is suitable for studying peptides purified from natural sources, and this will be the focus of Section 9.09.3. 

In a homonuclear decoupling experiment a particular multiplet is irradiated suppressing the coupling interaction between the irradiated nucleus and its coupling partners. A comparison of the standard coupled ID spectrum and the selectively homonuclear decoupled spectrum reveals which nuclei are coupled. Whether a homonuclear decoupling experiment or a 2D homonuclear COSY experiment would be the best solution for multiplet analysis in a one-dimensional spectrum depends very much on the nature of the problem under investigation. If a large number of multiplets need to be irradiated then a two-dimension approach may be preferable. 

An alternative approach to tailor the cross peaks is the z-filtered COSY spectrum (z-COSY) [5.130]. In Check it 5.4.1.7 the "small-flip angle COSY" spectrum of 2,3-dibromopropionic acid, the basic sequence of the z-COSY spectrum, is simulated. As such a z-COSY spectrum can not be calculated because the randomly changing delay which is the major part of the z-filter can not be simulated in the current version of NMR-SIM. A comparison of the results of Check it 5.4.1.7 and the E.COSY spectrum of the same spin system calculated in Check it 5.4.1.6 shows that due to the small flip angles the diagonal peaks of the z-COSY spectrum are reduced in intensity while the cross peaks are very similar. 

We have already discussed in Section 9.5.1 the type of information that NMR experiments can provide about the conformation of a molecule and the use of distance geometry for determining structures that are consistent with the experimental data. In the simplest molecular dynamics approach, we could incorporate harmonic restraint terms of the form k(d - dg) where d is the distance between the atoms in the current conformation and dg is the desired distance dynamics approach derived from the NMR spectrum, k is a force constant, the value of which determines how tightly the restraint should be applied. The information provided by the COSY experiment can also be expressed as a torsion angle via the Karplus equation torsional restraints may be incorporated into the molecular dynamics energy function as an alternative to the use of distances. There are many other ways in which the restraints can be incorporated for example, some practitioners prefer to penalise a structure only if the distance exceeds the target ... 

Over the following years many experiments were developed whereby the signal structure contains a lot of information on the coupling constants, cf. E-COSand P.E.COSY. Others stem from heteronuclear experiments, such as the method of Bermel et which is a good illustration of the way problems were approached at the time, and, more recently, the heteronuclear XLOC experiments which show structures similar to E.COSY spectra. 

---