Construction of theoretically (and practically!!) optimal
NMR pulse sequences
click to zoom
The design of a particular NMR experiment is dictated by
the type of spectral information desired. For complex spin
systems in large biomolecules, specific experiments may
be exceptionally difficult to construct when high sensitivity,
suppression of spurious signals, and tolerance to relaxation
effects are required. For example, multidimensional NMR
experiments for detection and quantification of conformational
exchange are usually tend to suffer from low sensitivity,
which can be improved by utilizing all available magnetization;
the 13C-labelled methyl groups represent one of the most
complicated systems of J-coupled homo- and heteronuclear
spins found in solvated proteins and nucleotides. The availability
of a general tool to construct NMR experiments correlating
arbitrary transitions in such a spin system with near-optimal
sensitivity is extremely useful. Such a complex optimization
problem in a space of high dimensionality turns out to
be numerically tractable. Based on the application of molecular
dynamics in the space of pulse-sequence variables, a general
method is proposed for constructing optimized coherence
transfer elements (CTE) capable of transferring an arbitrary
(generally non-Hermitian) spin operator encoding the chemical
shift of heteronuclear spins to an arbitrary spin operator
suitable for signal detection. The CTEs constructed in
this way are evaluated against benchmarks provided by the
theoretical unitary bound for coherence transfer and the
minimal required transfer time (when available). This approach
is applied for constructing of a set of NMR experiments
enabling direct and selective observation of individual
1H-transitions in 13C-labeled methyl spin systems close
to optimal sensitivity.
Automated backbone and side-chain assignment of biopolymers
from NMR data
click to zoom
There are an estimated 90,000 genes in the human genome
(ref), many of which encode several protein isotypes. Furthermore
there are countless human pathogens, each with their own
proteome. The goal of structural genomics is to determine
the structure of as many as possible of these proteins
of direct human interest. Considering the vast number of
proteins involved, it is clear that the current methods
for high resolution structure determination cannot meet
the demand for structural information, since the expected
time commitment for either NMR or crystallography ranges
from months to years for each structure determined. Some
change in paradigm must therefore be adopted in order to
reduce this time, and thus increase the throughput of structure
determination. In the case of NMR, the primary bottlenecks
in the structure determination process are data acquisition
and resonance assignment. we introduce a new program for
the assignment of sidechain resonances. Like AutoLink,
this new program, called SideLink, uses Relative Hypothesis
Prioritization to emulate “human” logic. In order to address
the higher complexity of sidechain assignment problems,
the RHP algorithm has itself been advanced, making it capable
of processing almost any combinatorial logic problem. Additionally,
SideLink directly examines spectral data, overcoming the
need and limitations of prior data interpretation by users.
Structural studies of Leu-Zippers
click to zoom
Are the Leu-Zippers, a protein dimerization domain occurring
in many eukaryotic enhancer-type transcription factors,
referred to as basic leucine zipper proteins (bZIP) and
basic helix-loop-helix leucine zipper proteins (b-HLH-LZ),
are really mere dimerization motifs? Leucine zipper transcription
factors have evolved as regulators in many processes that
are critical to the function of an organism, from cell
metabolism and differentiation to circadian rhythms, memory
and development of organs. These factors are wide-spread
among eukaryotes, with only human genome containing 56
genes encoding 53 proteins with bZIP motifs. We determined
the resonance assignment, secondary structure and dynamic
properties of a stable non coiled coil conformation of
the dimerization domain from yeast transcription activation
factor GCN4 (Leu zipper; LZGCN4). A new line of fully optimized
spin state exchange experiments, XYEX-TROSY, applied to
1H, 15N and 1HN?13C??moieties, established that in a broad
range of pH and buffer conditions the classical LZGCN4
coiled coil dimer is in a dynamic equilibrium with another
distinct conformation (denoted here as x-form) and enabled
complete assignment of the resonances stemming from the
x-form. The LZGCN4 x-form is generally less structured
in comparison with the classical GCN4-p1 coiled coil, but
still retains a structured a helical central core. The
implications for folding properties and biological significance
are discussed
The Cytoplasmic Domain of the Chloride Channel ClC-0.
Structural and Dynamic Characterization of the Disordered
Regions
click to zoom
Eukaryotic members of the ClC family of chloride channels
and transporters contain next to their transmembrane ion
transport domain cytoplasmic domains, which are believed
to be involved in the modulation of ClC function. In some
family members those regulatory domains contain next to
a well folded structured part, long sequence stretches
with low sequence complexity. These regions, a 96 residue
long linker connecting two structured sub-domains, and
46 residues on the C-teminus of the domain were found disordered
in a recent crystal structure of this domain in ClC-0.
Both regions have a large influence in modulation of channel
function in closely related family members. Here we describe
a NMR study to characterize the structural and dynamic
properties of these putatively unstructured domains. Our
study reveals that the two regions indeed show large conformational
flexibility with dynamics on the nanosecond timescale,
that, however small stretches of secondary structure are
found interdispersed between the unfolded regions. This
study characterizes for the first time the biophysical
properties of these protein segments, which may become
important for the understanding of ClC regulation.
A Minimal Transmembrane beta-barrel Unit Studied by
NMR Spectroscopy
click to zoom
Outer membrane protein A (OmpA) is one of the major outer
membrane proteins of Escherichia coli (E. coli). This protein
is composed of two domains: an N-terminal transmembrane
(TM) domain and a C-terminal periplasmic domain. The three-dimensional
(3D) structure of the TM domain has been determined by
X-ray crystallography as well as by nuclear magnetic resonance
(NMR) spectroscopy. This latter domain is composed of 8
antiparallel ?-strands connected by four relatively large
and hydrophilic surface-exposed extra-cellular loops and
three short periplasmic turns into a ?-barrel. In our studies
of outer membrane protein A (OmpA), we were concerned with
the structural role of the surface-exposed extra-cellular
loops of the N-terminal transmembrane domain of OmpA. An
OmpA variant with all four surface-exposed loops shortened,
which we call the beta-barrel platform (BBP), was successfully
refolded. This indicates that the removed parts of the
surface-exposed loops indeed do not contain amino acid
sequences critical for this membrane protein’s refolding
in vitro. BBP can therefore serve as a good starting point
for the development of integral membrane proteins with
novel engineered functions. We have determined the global
fold of BBP+EF, an OmpA variant with all four surface-exposed
loops shortened and a surface-exposed EF-hand loop metal
binding site inserted (in one of the shortened surface-exposed
loops), by solution nuclear magnetic resonance spectroscopy.
BBP and BBP+EF in DHPC micelles consist of 8-stranded antiparallel
?-barrels. Finally, BBP represents the smallest ?-structured
integral membrane protein known to date.
Structural plasticity of peptidyl-prolyl isomerase
sFkpA as a key factor to its chaperone function
click to zoom
FkpA is a heat shock periplasmic peptidyl-prolyl cis/trans
isomerase (PPIase) with chaperone activity. FkpA was originally
discovered as a periplasmic Escherichia coli homolog of
the Macrophage Infectivity Potentiator (MIP) protein-like
FK506-binding proteins. Overexpression of FkpA suppresses
the formation of inclusion bodies from a defective folding
variant of the maltose-binding protein and promotes the
reactivation of denatured citrate synthase. Coexpression
of FkpA can dramatically improve functional periplasmic
production of single-chain fragments (scFv) of antibodies,
even those not containing cis-prolines. The yield of soluble
and functional scFv fragment was also increased in vitro
in the presence of stoichiometric amounts of FkpA. Recently,
it was reported that covalent fusion with FkpA can increase
functional solubilization of aggregation-prone HIV envelope
proteins and that the catalytic parameters for hydrolysis
of ampicillin by scFv9G4H9 are clearly influenced by the
presence of FkpA. The folding-assisting function of FkpA
was previously hypothesized to be due to its interaction
with early folding intermediates preventing their aggregation,
and its ability to reactivate inactive proteins, possibly
by binding to partially unfolded species. Intramolecular
dynamics of periplasmic chaperone FkpA-?CT (sFkpA) and
its complexes with partially structured substrates are
studied by NMR in solution. The backbone amide 15N relaxation
of sFkpA reveals flexibility in the relative orientation
between the dimerization domain and two juxtaposed catalytic
domains identified in the X-ray structure of sFkpA. This
flexibility is attributed to the structural plasticity
within the long ??helical arm (helix III) consisting of
residues 84 and 91. Residual dipolar couplings (RDCs) indicate
an absence of fixed orientation between the sFkpA domains.
The substrate binding surface of sFkpA is defined on the
X-ray structure by mapping of chemical shift perturbations
introduced by complexation of sFkpA with its corresponding
protein substrates: partially folded RNase A S-protein
and reduced carboxymethylated bovine ?-lactalbumin (RCM-la).
A comparison of 15N relaxation of apo-sFkpA and its complex
with RNase A S-protein indicates an increased rigidity
within the long ??helix III and decreased interdomain mobility
of the complex. We speculate that these dynamic properties
may play a key role in the chaperone activity of sFkpA,
since ability to bind different substrates potentially
requires structural adaptations of the chaperone protein.
We show that binding of sFkpA to RNase A S-protein greatly
reduces the population of aggregated oligomeric species
of RNase A S-protein. Finally, a molecular model, the so-called “mother’s
arms” model is proposed to illustrate the mechanism of
chaperone activity by FkpA.
Taking structural plasticity of an enzyme to extreme
click to zoom
Although protein dynamics have been recognized as a potentially
important contributor to enzyme catalysis, structural disorder
is generally considered to be deleterious to catalytic
efficiency. This widely held assumption has recently been
challenged by the finding that an engineered chorismate
mutase combines high catalytic activity with the properties
of a molten globule, a loosely packed and highly dynamic
conformational ensemble. Taking advantage of the ordering
observed upon binding of a transition state analog, we
have now exploited NMR spectroscopy to determine the structure
and dynamics of this enzyme. The enzyme-ligand complex
adopts a helix bundle structure, as designed, but retains
unprecedented flexibility on the millisecond timescale
across its entire length. These motions coincide with the
independently measured rate of product release, suggesting
that they may gate ligand binding. Pre-steady state kinetics
further show that the transition between the molten state
and the highly dynamic complex occurs by an induced fit
mechanism on the same timescale as the enzymatic reaction,
thus linking global conformational plasticity and efficient
catalysis. The observation that a flexible conformational
ensemble can achieve rate accelerations identical to those
of a native-like enzyme scaffold contravenes conventional
wisdom and suggests that our view of efficient enzymes
must be expanded to include structurally disordered proteins.
Projects
