Isotope labeling of proteins

From NMR Wiki

Jump to: navigation, search


Motivation of labeling proteins

Proteins are big macromolecules that contains thousands of protons, the span of the proton chemical shift is around 12 ppm, that's mean around 1000 of Hz, so to make all resonating frequencies distinct we should have a signal that last for several seconds. And that's for from being the case in protein. So to make separation of the resonating frequencies of each protons, we need to make multidimensional experiments. The problem with proton is also that they couple only through at least 3 bonds ( to go from a proton that is attached on one heteroatome, to another one ), so this make longer preparation of the spin system, and this lead to signal worsening. For big protein, proton become less and less suitable, because it has a huge bipolar coupling to other atoms, and this lead to enhanced relaxation. This problem could be circumvented, partially by using other NMR experiments, but it could be not enough efficient. So in this case we could do labeling with Deuterium ( the dipolar interaction with a proton is 42 time smaller than that between two protons ).

Labeling could also allow to have only one part of a protein, visible. For instance to see only the DNA, and not the protein.

Different kinds of labeling

Most frequently, proteins are labeled with 15N and 13C. 2H labeling is used for proteins of larger molecular weight.

Sources of isotopes used for labeling

The simplest labeling, and also the cheapest one is 15N, because 15N ammonia is quite cheap. 13C is more expensive, because it requires the synthesis (most commonly - biosynthesis) of 13C glucose, some expression systems allow use of cheaper 13CO2.

Uniform labeling

In uniform labeling all atoms of a selected element are represented by a single isotope.

Partial labeling

This is done mostly for reduction of line widths in the spectra. Unfortunately it's not possible to use 15N of the amino acid to label because cell in which we express the protein have transaminase that make fast exchange of the label. Deuterium labeling could be done only for a portion of all hydrogens. The partial labeling is important because we want to still have NOE peaks in the NOESY. With 50 % 2H incorporation, we have only 25% of the atoms pairs to be coupled in NOE experiments. but the peak could be observed in greater range, and also deuterium incorporation make them less broaden, so that they overlap less, and relax slower.

Deuterium has also the disadvantage of causing the isotopic chemical shift, it's a change in the chemical shift caused by the exchange with another isotope of one neighboring atoms, so this could lead to more complicated spectra.

Site-specific labelling

In the site-specific labeling approach only certain residues, or particular atoms in some residues are isotopically labeled.

Site-specific labeling makes spectra simpler, and can help in stereospecific assignment of certain resonances. Site-specific labelling also can permit structure determination of the main chain alone.

One recent approach to site-specific labeling is the SAIL [1] labelling, it employs a set of labeled amino acids designed to prepare samples giving the most useful NMR information.

Synthesis of labeled proteins

Expression in minimal media

A number of protein expression systems can be used to express proteins in Minimal media


  1. Kainosho, M and Torizawa, T and Iwashita, Y and Terauchi, T and Ono, AM and Guentert, P. Optimal isotope labelling for NMR protein structure determinations. Nature 440(7080):52--57, 2006. BibTeX [kain06]
  2. MacKenzie, DA and Spencer, JA and Gal-Coeffet, MFL and Archer, DB. Efficient production from Aspergillus niger of a heterologous protein and an individual protein domain, heavy isotope-labelled, for structure-function analysis. Journal of Biotechnology 46(2):85--93, 1996. BibTeX [mackenzie1996epa]

Personal tools