SANDEEP KUMAR, Ph.D.
Contact Information
Laboratory of Experimental and Computational Biology,
National Cancer Institute (NCI) - Frederick,
Building 469, Room 151,
Frederick, MD 21702.
Phone: 301-846-6542;
Fax : 301-846-5598; Email:
kumarsan@ncifcrf.gov.
Present Position
I am a postdoctoral visiting fellow with Drs. Ruth Nussinov and Jacob V. Maizel
at LECB in NCI-Frederick.
Education
Ph.D. Computational Molecular Biophysics
I obtained my Ph.D. degree from
Molecular Biophysics Unit,
Indian Institute of Science,
Bangalore, India in 1998.
My doctoral research was focused on sequence structure relationship
in alpha helices.
Ph.D. thesis title: Geometry and sequence correlation studies on
alpha helices in globular proteins.
M.Sc. Biotechnology and Molecular Biology
In 1992, I obtained my Masters degree in Biotechnology
and Molecular Biology from G. B. Pant University.
My M.Sc. research focussed on Glutamine synthetase production by Bacillus brevis BbG1
under different physiological conditions.
M.Sc. thesis title:
Computerized mathematical approach for kinetics and production of Glutamine synthetase
by B. brevis BbG1.
B. Sc. (Hons.) Physics
I completed my undergraduate studies in Physics (B. Sc. (Honours) Physics)
from University of Delhi in 1989.
Research Interests
Computational molecular biophysics, sequence-structure-function relationship in
biological macromolecules, specially proteins. Protein folding, binding and
stability. Electrostatic interactions in protein stability. Relationship between
"macroscopic" thermodynamic properties and "microscopic" protein structural
properties. De novo protein design, bioinformatics and computational biology.
Description of present research
My present research explores physico-chemical rules behind protein
folding, binding and stability as well as relationships among sequence, structural and
thermodynamic data on proteins.
Protein thermostability
Recent years have witnessed an explosion in the available structural
information on biological macromolecules, especially proteins.
Such an information can be usefully exploited to study a number of
biochemically relevant problems like molecular adaptation of proteins
found in organisms living under extreme environmental conditions. I have
compared various structural and sequence properties, namely,
hydrophobicity, packing, oligomerization, insertion/deletions, hydrogen bonds,
salt bridges, amino acid compositions and residue substitutions for eighteen
non-redundant families of thermophilic and mesophilic proteins.
Correlation between electrostatic interactions and protein thermostability was
most consistent among the properties compared in this
study.
Further, I have compared electrostatic contributions of salt bridges towards
stability of hyperthermophilic Pyrococcus furiosus glutamate dehydrogenase
and mesophilic Clostridium symbiosum glutamate dehydrogenase monomers.
Salt bridges in Pyrococcus furiosus glutamate dehydrogenase form extensive
networks and, as a result, have greater electrostatic stabilities than those in
Clostridium symbiosum glutamate dehydrogenase. Increased number of salt
bridges and their networks contributes towards
thermostability of Pyrococcus furiosus glutamate dehydrogenase.
Role of electrostatic interactions in protein stability and flexibility
Electrostatic interactions have important roles in protein folding, binding
and stability. Electrostatic contribution of salt bridges can be stabilizing, insignificant
or destabilizing towards protein structure depending upon their geometry, location and
interaction with other charges in the protein globule. Recently, I have investigated
role of the salt bridges in protein stability
by computing
their electrostatic contributions in several monomeric proteins.
Salt bridges with 'good' geometry are usually stabilizing towards the proteins.
There are two types of protein flexibilities, viz., segmental
and systemic. Segmental flexibility refers to rigid body movement of two or more subparts
(domains, subunits, etc.) of a protein with respect to one another e. g. hinge-bending
motion. Segmental flexibility is usually occurs on slow time scale. Segmental flexibility can
be characterized in structural details by comparing "open" and "closed" conformations of
enzymes. Close range electrostatic interactions such as salt bridges and hydrogen bonds
are usually avoided across the moving parts of the protein. In contrast, systemic flexibility
refers to fast movements of protein backbone and side chain atoms about their mean positions.
Unlike the segmental flexibility, systemic flexibility is distributed all over the protein.
Systemic flexibility can be analyzed by using multiple low energy conformations around the
native state of the protein, such as those obtained from NMR experiments or by performing
molecular dynamic simulations.
I have analyzed the electrostatic contributions of six intra- and inter-helical
salt bridges/ion pairs and an ion pair network in 40 NMR conformers of c-Myc-Max leucine
zipper. The electrostatic contribution of each ion pair and the ion pair network fluctuates
in conformer dependent manner. Each ion pair and the ion pair network inter-converts
between being stabilizing and being destabilizing at least once in the 40 conformers
of the NMR ensemble. The origin of fluctuations in the electrostatic contributions of the
ion pairs can be traced to movements of the charged residues in proteins. This
study
indicates that overall contribution of salt bridges/ion pairs in solution may vary in
conformer-population dependent manner.
A large scale analysis of 22 ion pairs in 14 NMR conformer ensembles (with atleast
40 conformers), average energy minimized structures and multiple crystal structures of
11 non-homologous proteins has confirmed the above observations. These investigations have
resulted in improved understanding of systemic protein flexibility.
I have used the data from this analysis to characterize the relationship between
geometries and electrostatic strengths of the ion pairs. It appears that oppositely charged
residue pairs are mostly stabilizing towards proteins if their side chain charged group
centroids are within five angstrom distance. These observations are useful for identifying
stabilizing electrostatic interactions in protein structures and in de novo protein design.
Statistical analysis of protein thermodynamics data collected from experimental literature
There are two kinds of information available on proteins based on "macroscopic" and "microscopic"
properties of the proteins. Thermodynamic information such as enthalpy, entropy, heat capacity
and free energy differences between the folded and unfolding states of a given protein describes
the macroscopic properties. Such information is usually gathered by spectroscopic
(UV/VIS, CD, NMR), microcalorimetry (DSC) and hydrogen-deuterium exchange experiments. These
experiments typically involve protein denaturation using physical (temperature) and chemical
(Urea, GdHCl) agents. The data obtained using these experiments can be used to plot protein
stability curves described by Gibbs-Helmholtz equation. A protein stability curve describes
variation in the Gibbs free energy change between folded and unfolded state of the protein as
a function of temperature. The information on microscopic properties of a protein is obtained
from its atomic coordinates. The atomic coordinates can be obtained by solving the protein
structure using crystallography and NMR. The challenge is to relate the two. One way to approach
this problem is to analyze proteins for which both the thermodynamic data and the structural
information is available. Using this approach, I am trying to learn more about
protein thermostability. I have compared
the protein stability curves for the families containing homologous thermophilic and mesophilic
proteins. The observed differences are interpreted in terms of the sequence and structural
differences in these homologues. Recently, I have written a review article on protein
thermostability. This article tries to relate macroscopic thermodynamic differences in
proteins with the differences in their sequence and structural properties. The central
question asked in this article is:
How do thermophilic proteins deal with heat? The observations
on the thermodynamic differences among the homologous thermophilic and mesophilic
proteins indicate that the thermophilic proteins achieve greater temperature resistance
via increased thermodynamic stability. The protein stability curves of the thermophilic
proteins are up-shifted and broader than those of their mesophilic homologues as shown by
this
study.
Thermophilic proteins have greater enthalpy change at melting temperature, smaller heat
capacity change and greater maximal thermodynamic stability as compared to their mesophilic
homologues. Formation of specific interactions, such as electrostatics, may be responsible for
this observation. This is consistent with our previous studies. Another interesting
aspect of this analysis is that stabilities of the homologous thermophilic and mesophilic
proteins at respective living temperatures of the source organisms are similar.
Most proteins are maximally stable around the room temperature!
Hydrophobic effect is the major force driving protein folding. It
also explains clathrate formation by small organic solutes in water. Hydrophobic effect
is strongest around the room temperature. Consistently, small organic solutes and apolar
amino acids show minimal solubilities in water around the room temperature.
A recent
analysis of proteins that show reversible two-state folding/unfolding transition around
the neutral pH has revealed an interesting observation. Such proteins with sufficiently large
hydrophobic core are maximally stable around the room temperature, irrespective of their
amino acid sequences, structural folds and living temperature of the source organisms.
Critical building blocks, protein folding/misfolding and protein folding pathways
Proteins are made up of building blocks at several hierarchical levels.
Different building blocks have different contributions towards protein structure and stability.
One or more of these building blocks may be critical for correct protein folding.
In the absence of such critical building blocks, the proteins may misfold into
non-native conformations. I have developed an algorithm to identify such critical
building blocks. So far, I have studied the
folded structure of adenylate kinase using this concept. We find that the critical
building blocks also contain functionally important residues. This indicates that
protein folding and function may be coupled. Occurrence of residues important for both
protein folding and function in the same segment may be evolutionarily advantageous.
In such proteins, the nature may need to guard against mutations in a single segment
to protect both protein folding and function. Using a non-redundant data set of 930
protein chains with non-homologous structures, we have identified such critical
building blocks in 225 proteins. Most of these proteins fold in complex non-sequential
manner. Presently we studying structure-function conservation in these proteins.
Implications of energy landscape theory for protein folding and binding
A "new view" of protein folding is emerging from lattice modeling
studies. This view is based on statistical mechanical concepts and describes
protein folding as multiple pathway process. Different protein molecules are
thought take different paths down their energy landscapes to reach native conformation.
Some of these routes may be more frequently traveled than others.
I am participating in discussions aimed to explore implications
of this new view to folding and binding of large proteins.
Using this theory, the "Lock and Key" and "Induced fit" models for protein-protein,
protein-ligand, enzyme substrate binding can be understood in terms of
"conformer selection". This theory is also useful in understanding and interpreting
protein flexibility. These discussions are summarized in about half a dozen articles
on this subject from our group. These articles are listed in
publications.
Description of research for Ph.D. degree
Alpha helix is a major secondary structural element in proteins.
My Ph.D. research focused on statistical analysis of sequence and structural characteristics
of the alpha helices in globular proteins. It has been known for quite sometime that alpha
helices in globular proteins often curve, bend and kink. I had developed HELANAL,
a program to characterize alpha-helix geometries in proteins. This program classifies
geometry of an alpha helix as being linear, curved or kinked. My studies on alpha
helices in globular proteins showed a wide diversity in length, geometry and location
for these motifs. The origin of these variations can be traced back to the amino acid
sequences of the alpha helices. A detailed description of my observations is presented in
in this publication.
Different positions in alpha helices, e.g. helix N- and C- termini and middle,
prefer and avoid different amino acids. For example, Ncap position of protein alpha helices is very
exclusive. It prefers six amino acids and avoids eleven others. I have carried out
a comprehensive analysis of amino acid preferences and avoidances at 15 different positions in and
around alpha helices. Each of these position shows unique preference and avoidances for different amino acids.
Furthermore, there are several different structural motifs found at the alpha helix termini. The
presence of these motifs is correlated with the occurrence of particular amino acids at the helix
termini.
Something you may like to download from here:
This talk presents an interesting aspect of my work. It highlights the fundamental
nature of electrostatic interactions and their role in protein stability as well as in
molecular adaptation. This talk is for educational / informational purpose only and should not be construed as
a final word on this extremely complex issue. I expect you not to copy this talk (completely or in parts)
and pass it on as your own or somebody else's (other than mine) intellectual property. It is a powerpoint
presentation.
Often your results are as good as the data you collect. In protein thermodynamics and
kinetics of protein folding / binding this is especially true. The experimental thermodynamics data on
proteins reported from
different labs across the world is very heterogeneous. This situation is compounded by the observations
that the same protein may show two-state or three-state folding--unfolding transitions under different
conditions such as pH, buffer, salt concentration, presence / absence of ligands, cofactors, metal ions,
substrates or denaturants. Hence, one of the challenges that I faced was to devise
a gold standard for what should be called as a reversible two state protein. I arrived at this standard
after reading thousands of papers that report the calorimetric and spectroscopic experiments aimed at
determining the protein thermodynamic parameters. The above manuscript defines this gold standard and
the dataset that I obtained from the literature for the proteins that show a reversible two-state
folding -- unfolding transition at or near the neutral pH under the stated conditions. I would suggest
the use of these criteria and the dataset by experimentalists and theorists in their work.
This
is also the supplementary material for my 1999 JMB paper on salt bridge stability in proteins.
A few useful links
Last updated on November 18, 2002.