Protein Preparation
SAIL-labelled protein is prepared using cell-free methods. The amino acids used for the protein production are prepared using chemical and enzymatic syntheses. Their labelling is guided by the following principles:
– in each methylene group, one of the ¹H atoms is stereo-selectively replaced by a ²H atom.
– in each methyl group, two of the ¹H atoms are replaced by ²H atoms
– the prochiral methyl groups of Leu and Val are stereo-selectively ¹²C²H₃ and ¹³C¹H²H₂ labelled
– six-membered aromatic rings have alternating ¹²C²H and ¹³C¹H moieties
The SAIL Wiki site has a figure illustrating the labelling scheme.

Applications
The SAIL labeling pattern is particularly useful for the structure calculation of large proteins with advantages seen in the NOESY spectra. Non-exchangeable side-chain protons are essential for defining side-chain conformations, but are often prone to overlap. By reducing the number of these protons in SAIL-labelled samples, the overlap is decreased and the quality of NOESY spectra is improved. Indeed, although there are fewer NOE peaks for a SAIL-labelled protein than for a uniformly ¹⁵N,¹³C-labelled one, it is likely that there are more interpretable NOE peaks. In any case, many of the peaks which are removed, have a low information content because they involve fixed (geminal) distances or duplicate information already present from other NOEs. In addition, the information obtained is immediately stereo-specific. A further advantage is that the most serious sources of spin diffusion are removed which improves the accuracy of inter-proton distance measurements.
More general advantages of SAIL-labelled samples include the fact that lines become sharper due to a decrease in long-range couplings and by eliminating dipolar relaxation pathways. The aromatic ring labeling strategy removes one-bond ¹³C-¹³C couplings, which often complicate spectra or require the use of constant time data collection methods to reduce spectral complexity. And the measurements of couplings is made easier because CH₂ and CH₃ groups are converted to CHD and CHD₂ groups.
The labelleding of methyl and methylene groups in SAIL-labelled samples also makes them amenable to straight forward relaxation experiments for the investigation of side-chain motions.
Despite the clear advantages to using SAIL-labelled protein for structure determination, this labelling method is not very common – probably because it is very expensive and requires cell-free technologoy which is not available in many labs.

Reference:
M. Kainosho, T. Torizawa, Y. Iwashita, T. Terauchi, A.M. Ono and P. Güntert (2006) Nature 440 52-57. (Link to Article)

SAIL Wiki run by the Kainosho and Güntert labs