 |
Determination of the three dimensional Structure and HN--S Hydrogen Bonding of the Synthetic 113Cd3bN domain of
Lobster MT-1 by NMR
Amalia Muñoz#, F. Holger Försterling, C. Frank Shaw III+ and David Petering
Department of Chemistry and UWM-NIEHS Marine and Freshwater Biomedical Core Center,
University of Wisconsin-Milwaukee, Milwaukee WI 53211
# Current address: Dept. of Biochemistry, University of Wisconsin-Madison, Madison, WI 53706
+ Current address: Department of Chemistry, Eastern Kentucky, Richmond KY-40475 |
|
Abstract
Metallothionein (MT) is a cysteine-rich metal-binding protein. MT clusters
of different origins, exhibiting a range of highly
conserved cysteine positions in their sequences, show differences
in metal-cysteine connectivities and reactivity.
Lobster-MT, which includes two Cd3S9 ß-domains,
was chosen as a basic model to study the structure-function
relationship among clusters. Here NMR studies concerning the ßN
domain will be presented. The metal clusters were
defined from the 2D 1H/113Cd HSQC-TOCSY experiments
and found to exhibit the same metal-cysteine coordination as
in the native holo-Cd6-MT from lobster and crab. The 1H/113Cd
HSQC-TOCSY experiment allowed us, from the amide proton region, to easily
identify the bridging and terminal cysteines coordinated to each cadmium.
Proton resonances were assigned using two dimensional methods; the
structure calculation was performed using 271 proton-proton distance restraints
from NOE data, and 37 restraints for the f and
Cys-c1 dihedral angles derived from 3JHNHa
and 3JHa/Hb.
In addition 6 restraints for the Cys-Cd c2 angle
derived from heteronuclear 1H/113Cd couplings were
included. Four possible HN-S hydrogen bonds were defined using
heteronuclear spin echo difference experiments, the strongest of which
was found to involve Glu-10. The NMR results also indicated the presence
of a second minor conformation with different metal environment as shown
by the differences in the 113Cd- and 1H-NMR chemical
shifts. The structure and implications of the hydrogen bonding pattern
for sequence preservation and possibly in the conformational changes will
be discussed. The results shown here indicate than the ßN-domain
has a
more labile structure than the ßC. These differences can be
related to the differences in reactivity observed between these
two Cd3ß domains toward EDTA and DTNB (5,5'-dithio-bis-(2nitrobenzoic
acid)). Therefore, this work shows the
correlation between cluster structure and reactivity among the ß-domains
in metallothioneins.
Introduction
Metallothionein (MT) is characterized by its high content in cysteine residues (30%) and lacking aromatic amino acid and histidines. Metallothionein also lacks disulfide bonds,
thus is capable of bind a large number or metal ions (7 in the case of mammalian MT). The first function attributed to MT was metal detoxification (Hg2+, Cd2+
), then roles as metallo-regulator of Zn2+ and Cu+ followed as well as the detoxification of reactive oxygen species and metabolism of metallodrugs
and alkylating agents.
Human and mammalian MTs bind 7 Zn2+ or Cd2+ ions via 20 cysteine residues, which are distributed in two independent and kinetically and thermodynamic
ally different clusters, Cd4S11 and Cd3S9, located at the a- and bbeta;-domains of the protein, respectively.
Crustacean MTs contain only 18 cysteines and 6 Cd2+, which generate two Cdi3S9 clusters. Previous structural studies carried out in both
species by 2D-NMR techniques show that the Cd-cysteine coordination and therefore the folding differs among these Cdi3S9 clusters in the bbeta;-domains
of crustacean and mammalian MTs. [1,2]
Previous studies showed that the reaction of lobster MT with thiol reagents such as DTNB (5,5'-dithio-2,2'dinitro benzoic acid) or DTP (dithiopyridine) were also biphasic but
more rapid than in the case of the mammalian MT reactions. Thus, the difference in cluster structure as reason for the biphasic kinetic was ruled out. And , it was proposed then,
that the kinetic and thermodynamic reactivity of the MT domains is related to the detailed folding of the peptides around their clusters.
To analyze structure-function relationships among these domains, in this poster we describe the 3D-structure of the chemically synthesized native lobster bbeta;
N (Figure 1) as determined by 2D-NMR spectroscopy.
Assignment and Secondary Structure:
Synthetic Cd3MT- bN exhibits 113Cd chemical shifts resembling cadmium ions in a tetrahedral environment, coordinated by terminal and bridging
cysteines. These shifts exhibit initially a time dependence, with the shifts approaching, but not completely reaching the shifts of the complete holo protein (Figure 2).
This is indicative of conformational rearrangement occurring. A stable structure obtained after three months was used for the structure calculations below.
Assignment of most proton resonances was obtained for the synthetic MT-bN using 2D TOCSY and NOESY experiments. The Cd-cysteine connectivities of the
Cd3S3i cluster were obtained from heteronuclear 2D- 1H{113Cd}HSQC-TOCSY experiments (Figure 3A). The results indicate that the
synthetic Cd3bN exhibits the same metal-coordination as the native holo protein (Figure 5).
Figure 3: The HN region of the 2D 1H{113Cd} HSQC-TOCSY spectrum (T=298K) shows correlations due to Sg-Cd bonds involving the respective Cys residue and allows one to
establish Cd-S cluster connectivities.
A total of 254 NOE's were obtained from 2D NOESY spectra obtained at 278K with mixing times of 60ms and 150ms. Measurement of the coupling constants 3JHNHa
, 3JHaHb and
3JHbCd yielded 13 additional dihedral angle restraints restraints for the backbone angle f, 10 for the side chain dihedral angle c1 and 6 for the CaCbbeta;SCd
dihedral angle, c2.
Analysis of the sequential to medium range NOE's and of 3JHNHa coupling constants indicated the absence of elements of secondary structure as expected for MT (Figure 4).
In addition, small and large values of the temperature coefficient of the HN chemical shift are also shown.
 |
Figure 4: Summary of sequential and medium range NOE's, 3JHNHa and temperature coefficients of d HN for Cd3-bbeta;
N. NOE's are classified as strong, medium, weak and very
weak indicated by different heights of the connecting box. Large Temperature coefficients (< -5 ppb K-1) indicating solvent exposed protons are drawn in red, whereas values
around 0 ppb K-1 indicating slow exchange are colored in green. A series of fast exchanging protons in the range of C16-G19 indicates increased flexibility of that loop.
|
Hydrogen Bonding
While temperature dependence of the amide protons (Dd > -5 ppb K-1provides an indication of possible hydrogen
bonds, scalar spin-spin couplings 2hJHNCd, normally in
the order of 0.5-2.0 Hz [5] and not observed in standard experiments, can provide direct evidence of a NH--S hydrogen bond. In the case of the
synthetic Cd3-MT-bN, a two dimensional 1H{113Cd}-HMQC experiment optimized
for small couplings and with decoupling of 3JNHaduring the indirect detection period (Figure 6B) was used
to obtain qualitative through hydrogen bond connectivity information between the HN and Cd atoms (Figure 6C). Additionally, the
magnitude of2hJHNCd was obtained from a one dimensional 1H{113Cd}HSED experiment [5] (Figure 6A).
This is not the first observation of a NH--S hydrogen bond by NMR, but it is the first example of a 2D HN-Cd correlation through a hydrogen bond
and the first example of an experimentally established hydrogen bond for a metallothionein in solution.
 |
 |
Figure 5: (A)Pulse sequence for 1D 1H{113Cd}
HSED experiment and (B) 2D long range1HN{113Cd} HMQC employing
a jump and return pulse pair for selective decoupling of 3JHNHaduring
t1 and water suppression.
f1 = x,y,-x,-y; f2
= -x, -y, x, y; f3 = 4 (-x), 4 x
f3' = 4 (-x) y =
2 (x, -x), 2 (-x, x)
D = 1 / [2 (dHN-dH2O)]
(208 ms) t = 1
/ (2hJHN-Cd) (62
ms)
(C) Long range optimized
1HN{113Cd}-HMQC exhibits cross peaks
indicative of either 2hJHN--Cd couplings (S24, C22,
E10) or long range intra residual 5JHNCd couplings
(C5). In addition, a cross peak involving a minor conformational
isomer is observed, presumably due to a hydrogen bond interaction.
Coupling constants obtained by a 1D HSED experiment are given in parentheses.
Structure Calculation
Structure calculations were carried out with the program XPLOR [3] employing an ab initio simulated annealing protocol. A subset of 300 structures was generated,
and the 30 structures of lowest energy were selected for the final analysis. Superposition of those 30 structures yielded a total rmsd of 0.52 A for the backbone, metal cluster
and cysteine-Cb atoms of the non terminal residues (residue 4-27), (1.24 A for all non terminal heavy atoms). For each of the two conformational isomers, the rmsd values of
the non terminal backbone and metal cluster atoms were 0.26 and 0.12 A, respectively.
 |
Figure 6: Ribbon
display of the synthetic MT-bbeta;N.
The Cd
-S
cluster and Cys- side chain are shown as ball and sticks. The numbering of the Cd atoms ia the same as in the holo protein of lobster-MT. Experimentally established HN--S hydrogen bonds are shown as dashed lines with donor residues indicated.
This figure was prepared using the program MOLMOL2k.1 by Koradi et al. [4]
Click on the picture to view it full size
|
Discussion of Structure
A ribbon representation of the minimized average structure is shown in Figure 6. The Cd-S cluster and the Cys side chains are also shown. The Cd-S cluster exhibits a boat
conformation as it is typically observed for b domains of MT. The superposition of the 30 structures of lowest energy (Figure 7) exhibits the
presence of two conformations in the loop between Ala12 ans Gly19. Many residues in that area also exhibit rapid amide proton exchange (Figure 4). This is also represented
by the backbone dihedral angles f and y which exhibit two ranges for residues 14, 15 and 17.
 |
 |
Figure 7: Overlay of the 30 structures of lowest energy of Lobster MT-bN and compared with the
mean structure of the N-terminal domain of Crab-MT [1].
|
Figure 8: Backbone dihedral angles (f and y) for the 30 stuctures of lowest energy of lobster Cd3-
bN. The two conformations are indicated ( o and o). For comparison,
the values for Crab-MT are also shown (o).
|
Comparison with Crab MT
In a comparison with Crab-MT it is noticeable that the HN chemical shifts of Cys5 and E10 differ significantly for lobster-MT when compared to Crab-MT (Figure 9). Since the difference is common to both the isolated domain Cd3-bN and the holo protein of lobster-MT, it has to be related to a structural change related to a change in the sequence. It is interesting to note that in the case of Cd3- b N a strong ross peak due to a NH--S hydrogen bond is observed between those two residues. Furthermore, E10 is substituted by V10 in the case of Crab-MT, giving rise to the speculation that the nature of residue 10 influences the strength of the hydrogen bond (E10)NH--Sg(C5) observed in MT-bN. Indeed a closer inspection of the two structures indicated the possibility of a hydrogen bond (G15)NH--O(V10) in the case of Crab-MT which is not possible if V10 is substituted by E10 in the case of lobster-MT. An overlay of the backbone and Cd-S clusters of Lobster-bN and Crab-MT is shown in Figure 10. With the general fold of the two backbones quite similar, there are nevertheless distinct differences in the cluster
 |
 |
Figure 9: HN proton chemical shift difference between the synthetic MT-bbeta;Ndomain and the holo proteins of
Lobster and Crab. Residues exhibiting major changes are indicated. The change for C27 is unique for the i
isolated bbeta;N domain, whereas the difference for C5 and E10 is due to a difference in sequence between lobster and crab-MT.
|
Figure 10: Overlay of the backbone of Lobster-bN and
Crab-MT[1]. |
Literature:
[1] Narula, S. S; Brouwer, M.; Hua, Y.; Armitage, I. M.; Biochemistry1995, 34, 620-631.
[2] Zhu, Z.; DeRose, E. F.; Mullen, G. P.; Petering, D. H.; Shaw, C. F. III; Biochemistry 1994, 33,
8858-8865.
[3] Brünger, A. T.; "X-PLOR: A system for X-Ray Crystollagraphy and NMR" , Yale University
Press, New Haven, CT, 1992.
[4] Koradi, R.; Billeter, M; Wüthrich, K.; J. Mol. Graph. 1996, 14, 51-55.
[5] Blarke, P. R.; Lee, B.; Summers, M. F.; Adams, M. W. W.; Park, J. B.; Zhou, Z. H.; Bax, A.; J.
Biomol. NMR 1992, 2, 527-533.
This page was created by F. Holger Försterling on May 11 2000
Last update [an error occurred while processing this directive]
