Protein Structure and Dynamics by NMR in Solution
105
4
Selected Applications
In what follows, we would like to illustrate the diversity of information gained
from NMR studies on proteins by the presentation of two recent research projects
of our laboratory. The first one will give an example of a classical structure determination of a protein module and the use of the chemical shift information for
the characterization of ligand-binding. The second one focuses on the obtention
of dynamical information on a ISN-labelled protein and the determination of the
tensor describing its anisotropic overall tumbling.
4.1
Structure Determination and Ca 2 + Binding of the EGF-Like Module
of the Human Complement Protease C1 r
C1, a multimolecular protease, triggers the classical pathway of complement
(Arlaud et al. 1987). Its activation and catalytic activity are mediated by two
homologous serine proteases, C1r and CIs, which form a Ca 2 +-dependent tetramer (C1s-C1r-C1r-C1s). These two proteases are modular proteins with the following modular composition: CUB-EGF-CUB-CCP-CCP-serine protease. The
structural determinants for the tetrameric organization have been shown to be
localized in the N-terminal part of these proteins (Thielens et al. 1990). Both C1r
and CIs EGF modules contain the typical consensus sequence for Ca 2 + binding
found in many EGF-like modules, and were therefore supposed to playa key role
in the intermolecular interaction. In addition, the EGF-like module of C1r (C1rEGF hereafter) possesses an unusually large loop between the first two cysteines
comprising 14 residues as opposed to 2-7 residues in other EGF-like modules.
We were interested to determine whether this large loop adopts a well defined
structure that could give insight into a possible functional role. Therefore, an
NMR investigation was performed in order to determine the molecular structure
of C1r-EGF and to study the structural consequences of Ca 2 + binding. A more
detailed description of this study can be found in Bersch et al. 1998.
4.1.1
Structure Determination of apo C1 r-EGF
C1r-EGF (C1r residues 123-175) has been synthesized chemically (Hernandez et
al. 1997). The first step of an NMR structure analysis consists of the assignment
of resonances eH resonances in this case) corresponding to the individual residues. According to a well-established procedure for non-labelled proteins (see
Basus 1989 for a detailed description), the intraresidual assignment is achieved
by using scalar information by means of COSY or TOCSY -like experiments.
Then, the residues are arranged according to the protein sequence using sequential NOESY information. Fig. 7.2 shows the fingerprint region of the NOESY spectrum in which each cross-peak corresponds to a through-space interaction
between two protons (H N and H U in F1 and F2 respectively) being less than 6.A
apart. From NOESY spectra acquired at different temperatures, 543 nOe could be
105
4
Selected Applications
In what follows, we would like to illustrate the diversity of information gained
from NMR studies on proteins by the presentation of two recent research projects
of our laboratory. The first one will give an example of a classical structure determination of a protein module and the use of the chemical shift information for
the characterization of ligand-binding. The second one focuses on the obtention
of dynamical information on a ISN-labelled protein and the determination of the
tensor describing its anisotropic overall tumbling.
4.1
Structure Determination and Ca 2 + Binding of the EGF-Like Module
of the Human Complement Protease C1 r
C1, a multimolecular protease, triggers the classical pathway of complement
(Arlaud et al. 1987). Its activation and catalytic activity are mediated by two
homologous serine proteases, C1r and CIs, which form a Ca 2 +-dependent tetramer (C1s-C1r-C1r-C1s). These two proteases are modular proteins with the following modular composition: CUB-EGF-CUB-CCP-CCP-serine protease. The
structural determinants for the tetrameric organization have been shown to be
localized in the N-terminal part of these proteins (Thielens et al. 1990). Both C1r
and CIs EGF modules contain the typical consensus sequence for Ca 2 + binding
found in many EGF-like modules, and were therefore supposed to playa key role
in the intermolecular interaction. In addition, the EGF-like module of C1r (C1rEGF hereafter) possesses an unusually large loop between the first two cysteines
comprising 14 residues as opposed to 2-7 residues in other EGF-like modules.
We were interested to determine whether this large loop adopts a well defined
structure that could give insight into a possible functional role. Therefore, an
NMR investigation was performed in order to determine the molecular structure
of C1r-EGF and to study the structural consequences of Ca 2 + binding. A more
detailed description of this study can be found in Bersch et al. 1998.
4.1.1
Structure Determination of apo C1 r-EGF
C1r-EGF (C1r residues 123-175) has been synthesized chemically (Hernandez et
al. 1997). The first step of an NMR structure analysis consists of the assignment
of resonances eH resonances in this case) corresponding to the individual residues. According to a well-established procedure for non-labelled proteins (see
Basus 1989 for a detailed description), the intraresidual assignment is achieved
by using scalar information by means of COSY or TOCSY -like experiments.
Then, the residues are arranged according to the protein sequence using sequential NOESY information. Fig. 7.2 shows the fingerprint region of the NOESY spectrum in which each cross-peak corresponds to a through-space interaction
between two protons (H N and H U in F1 and F2 respectively) being less than 6.A
apart. From NOESY spectra acquired at different temperatures, 543 nOe could be