On going Research Projects (In Brief):

Our focus is mainly on

– Development of NMR methodologies (Rapid data collection).

– Development of algorithms for NMR assignments.

– Structure-based functional genomics.

– Structure, dynamics, interaction of biologically important proteins and to correlate with their function.

– Structural and dynamic basis for understanding the way a protein sequence translates into a folded conformation.

– Effect of intracellular environment on the protein structure and dynamics.

– DNA structures in specific recognition with proteins.

In Detail:

Rapid Data Acquisition Methods in Protein NMR

Rapid Measurement of ³J(H^N-H^a) and ³J(N-H^b) coupling constants in polypeptides

We proposed two NMR experiments for rapid and accurate measurement of ³J(H^N-H^a) and ³J(N-H^b) coupling constants in polypeptides based on the principle of G-matrix Fourier Transform NMR spectroscopy and quantitative J-correlation. Rapid acquisition of 3D spectral data is accomplished by joint sampling of ¹⁵N and ¹H chemical shifts along one of the two indirect dimensions, which eventually results in a 2D spectrum. These experiments, namely, (3,2)D HNHA and (3,2)D HNHB facilitate acquisition of data with high spectral/digital resolution and chemical shift dispersion, overcoming several of the limitations in the currently used approaches. The experiments were demonstrated with two proteins of ~10 and ~15 kDa, representing a predominantly b-sheet and a-helical protein, respectively. Accuracy of the measured couplings was assessed by comparing it with those predicted from their respective NMR derived 3D structures. Rapid measurement of these coupling constants provided renewed opportunities to utilize them for sequence specific resonance assignments, estimation/characterization of secondary structure with/without prior knowledge of resonance assignments, stereospecific assignment of prochiral groups and 3D structure determination, refinement and validation. Taken together, these experiments will have a wide range of applications in structural genomics projects and to studying structure and folding in polypeptides. (J Biomol NMR. 2007 Dec;39(4):259-63)

Rapid Measurement of Pseudo-contact shifts in Metallo-proteins using a suite of GFT-experiments

Pseudocontact shifts (PCSs) provide valuable structural and dynamic information in (large) proteins. We proposed a methodology based on the principle of G-matrix Fourier Transform (GFT) NMR spectroscopy for simultaneous and rapid measurement of a large number of PCSs in proteins with a paramagnetic centre. Four experiments, namely, (3,2)D HNNCO, (3,2)D HNN(CO)CA, (3,2)D HNN(COCA)CB and (3,2)D HNHA taken together facilitate accurate measurement of six PCSs corresponding to ¹H^N, ¹H^a, ¹³C^a, 13C^b, ¹³C' and ¹⁵N nuclei. In addition, we proposed a new algorithm for unambiguous sequence specific resonance assignments of peaks shifted due to PCS. This avoids the need to record multiple 3D correlation experiments. The utility of the proposed experiments has been demonstrated with an 8.5 kDa protein, Calbindin. This provides new avenues to a wide range of applications in structure determination/refinement/verification and dynamical studies of proteins in general and paramagnetic proteins in particular. (http://www.bentham-open.org/pages/gen.php?file=16TOMRJ.pdf)

Methods for Identification of Spectral Signatures from their Immediate C-terminal Neighbours of Ala, Cys and Ser Residues in Uniformly ¹³C/ ¹⁵N-Doubly labelled proteins using GFT-NMR.

In an attempt for spectral simplification, we developed a pulse sequence, which enables directly to identify the spectral signatures arising from the immediate C-terminal neighbours of Ala, Cys and Ser residues in uniformly ¹³C/ ¹⁵N- doubly labelled proteins using GFT-NMR. Such assignments aid in sequence specific resonance assignment of any given protein. (Submitted)

Rapid stereospecific assignment of methyl groups using GFT.

In protein NMR, the focus during the last two decades has been on devising experiments to derive different types of structural restraints for elucidating their 3D structures. These constraints include distance and torsion angle constraints, chemical shifts, dipolar couplings etcetera. One such class of restraints is derived from stereospecific assignments of diastereotopic methyl (CH₃) groups of Val and Leu residues in proteins. These assignments provide the most important information concerning the orientation of isopropyl groups in Val and Leu residues about the C^a-C^b and C^b-C^g bonds, respectively. As a result, they have significant influence on the precision of derived 3D structures. To date, the most widely used approach for stereospecific assignment of CH₃ groups in Val and Leu residues relies on fractional ¹³C labeling. (Neri et al., 1989). However, severe spectral overlap in the case of large sized proteins (Mr > 20 kDa) poses a formidable problem. In particular, the cross peaks arising from Ala(C^b-H^b), Lys(C^g-H^g), Ile(C^g1-H^g1 and C^g2-H^g2), Met(C^e-H^e) and Thr(C^g-H^g) correlations have significant overlap with those of Val(C^g-H^g) or/and Leu(C^d-H^d) correlations, hampering the stereospecific assignments of the latter. Further, about 20 percent of the Val(C^g-H^g) cross peaks are seen to overlap with Leu(C^d-H^d) correlations. These overlap problems have been previously overcome by us using a methodology in which specific amino acids in a given protein are selectively ‘unlabeled’ while simultaneously allowing the fractional ¹³C labeling of other residues.

As an alternate to this we looked at the feasibility of modulating the chemical-shifts of the ¹³C nucleus directly attached to various ¹³CH₃ groups present in a given protein, using the GFT NMR principle. This simple trick propelled very significant dispersion in the spectral regions containing ¹³C-¹H cross peaks of CH₃ groups in Val and Leu residues with those of other amino acid residues. This lead us to carry out unambiguous stereospecific resonance assignments of Leu and Val methyl groups in partially ¹³C -labelled proteins. This also enabled us to develop a method to resolve nOe interactions with various methyl groups in ¹³C/¹⁵N-edited spectra of uniformly ¹³C labelled and ¹³C/ ¹⁵N- doubly labelled proteins. (Submitted)

Structure Based Functional Genomics.

The recent revelation of several complete genome sequences and large numbers of protein-coding regions have changed the whole course of structural biology. So far we have focused mainly on the details of protein function, addressing the question at the end: “How does the protein achieve its function?”. Recently we attempted to use protein structure to answer the question: “What is the function of a given protein?” In this endeavour, we have taken up structural characterization of two proteins (in collaboration with Dr. Yogendra Sharma, CCMB, Hyderabad). The first one is an 85 amino acid residue region of a protein annotated as bg-crystallin family protein from Methanosarcina acetivorans (an archaea). We have named this protein as M-crystallin. The other one is a 92 amino acid residue long protein from Hahella chenjuesis (a sea organism), which is supposed to have a single bg-crystallin domain with Greek key motif. We have named this protein as Hahellin.

In Search Of Ancestral bg-Crystallins in Archaea: Studies on a Putative Protein (M-Crystallin) having Sequence Signatures of bg-Crystallins from the Genome of an Archaea.

The evolution of complex and physiologically remarkable structures, like the vertebrate eye lens, has intrigued biologists for years. The transparent mammalian eye lens is made up of proteins called crystallins, which are classified into two distinct superfamilies called a-crystallins and bg-crystallins. While a-crystallin, a member of the small heat-shock protein family, functions as a molecular chaperone, no specific function has been assigned to lens bg-crystallins. However, b- and g-crystallins are shown to bind Ca²⁺ and thus they have been suggested to be involved in Ca²⁺ homeostasis in the eye-lens. Since b- and g-crystallins share similar structural and domain topology, they have been grouped together as bg-crystallins. Later, when some proteins were identified to have similar domain topology, they were grouped under bg-crystallins superfamily.

The bg-crystallin superfamily is sparsely distributed and includes highly diverse proteins found both in microbes and in higher organisms. Based on the structural topology, a few proteins were identified as members of this superfamily, although they do not show any sequence similarity. To date, only a few non-lens proteins which are closely related to lens bg-crystallins possess both the sequence conservation as well as 3D structural topology like Protein S from the bacterium Myxococcus xanthus, spherulin 3a from the slime mold Physarum polycephalum, and ciona crystallin from the urochordate Ciona intestinalis. Though these proteins have structural topology similar to that of lens bg-crystallins, there are distinct differences in their 3D structures as well as in their biophysical properties, such as Ca²⁺-binding, when compared with vertebrate lens bg-crystallins. Based on sequence homology, it has been suggested that these proteins might have an ancestral relationship with the lens bg-crystallins. Though protein evolution is thought to take place at a domain level, identification of a primitive and primordial bg-crystallin domains with similar structural homology to that of lens bg-crystallins remained elusive.

To date, bg-crystallins have been identified in two of the three domains of life on earth, namely bacteria and eukaryota. However, no protein with bg-crystallin domain has been identified yet in archaea. Based on their 16S rRNA sequence, archaea are among the oldest organisms on this planet. Hence, it would be evolutionarily interesting to study archaeal proteins. Since archaeal proteins are considered closer to those from eukaryotes than eubacteria, they might be precursors of homologous proteins in higher eukaryotes. Therefore, studying an archaeal homologue of bg-crystallins would be interesting not only to identify the presence of bg-domains in all kingdoms of life but also to establish an important missing link in the evolution of lens bg-crystallins and their Ca²⁺-binding properties. During the last decade, genomes of various archaea have been sequenced paving the way for such a study. On the evolutionary front, the exciting development was the sequencing and analysis of the genomes of several methanogens such as Methanosarcina acetivorans and Methanosaeta thermophillus. Further, methanogenesis seems to have appeared early in the evolution of Euryarchaeota. It is therefore, interesting to identify whether any protein with a bg-crystallin domain exists in a methanogen.

We searched for the presence of bg-crystallin homologues in the sequenced genomes of euryarchaeota and methanogens. We identified a putative protein having sequence signatures of bg-crystallins in the genome of Methanosarcina acetivorans, named as M-crystallin. This archaea is a mesophilic methane-producing anerobic archaeabacterium.

We determined the solution structure of M-crystallin in its Ca²⁺-free (apo) and Ca²⁺-bound (holo) forms using NMR and explore similarities with other members of bg-crystallin superfamily, and provide a vital clue regarding the evolution and origin of lens bg-crystallins.

In conclusion, we unraveled the solution structure of M-crystallin, the first archaeal bg-crystallin protein, in its Ca²⁺-free (apo) and Ca²⁺-bound (holo) forms using NMR spectroscopy and explored similarities with other members of bg-crystallin superfamily. The structure of this archaeal protein is strikingly similar to vertebrate lens b- and g-crystallins. To date, no other non-lens protein studied so far has such a close structural similarity with lens proteins. Additionally, M-crystallin exhibits Ca²⁺-binding properties (characterized by NMR, ITC, CD and fluorescence data) which are more similar to vertebrate lens proteins than microbial homologues. Such structural and functional similarities of M-crystallin in concert with the phylogenetic analysis strongly suggest that this primordial bg-crystallin domain protein from archaea is a possible ancestor of both amphibian and vertebrate lens bg-crystallins, thus providing a vital clue regarding the evolution and origin of lens bg-crystallins. (Submitted)

Structural Characterization of a protein from Hahella chenjuesis (a sea organism), which is supposed to have a single bg-crystallin domain with Greek key motif.

Doubly (¹³C and ¹⁵N) labelled Hahellin has been used for complete sequence specific ¹H, ¹³C and ¹⁵N resonance assignments in it. We are in the process to identify more proteins from this superfamily and carry out NMR structural studies.

Structure, dynamics, interaction of biologically important proteins and to correlate with their function.

Calcium binding proteins from Entamoeba Histolytica.

Entamoeba histolytica is a protozoan that is believed to be the causative agent for amoebiasis. It has been revealed that various calcium binding proteins (CaBPs) play a crucial role in its pathogenesis. The genome of Entamoeba histolytica revealed the presence of more than a score of CaBPs in this protozoan. Among these, we have recently determined the 3D structures of two EhCaBPs, namely EhCaBP1 and EhCaBP2. Based on both structural as well as functional studies, EhCaBP1 and EhCaBP2 are found to mimic the role of calmodulin in the protozoan. The fact that calmodulin is known to be present in other eukaryotics, it becomes very important to find out that whether calmodulin is also present in Entamoeba histolytica? If yes, what is its function? In this endeavor we have isolated, cloned and purified yet another calmodulin like protein (with 151 amino acid residue long; M^r = 17 kDa) and called as EhCaBP3, in collaboration with Prof. Alok Bhattacharya (JNU, New Delhi). Initial study of this protein reveals high degree of mutations in its conserved Ca²⁺-binding loops, which might make it inactive in Ca²⁺ signaling processes. Hence structural and functional characterization of EhCaBP3 is very important. In this endeavour, we have standardized the purification of the protein after varying several conditions in the various steps of purification. In continuation of these studies, we recorded a suite of double and triple [¹H, ¹³C and ¹⁵N] resonance multidimensional NMR experiments. Resonance assignment of the spectra is in progress. Doubly (¹³C and ¹⁵N) labelled EhCaBP3 has been used for complete sequence specific ¹H, ¹³C and ¹⁵N resonance assignments in it. We are in the process to identify more proteins from this protozoan and carry out NMR structural studies.

Energetics of Native Energy Landscape of a Two-domain Calcium Sensor Protein: Distinct Folding Features of the Two Domains.

The protein folding energy landscape provides a thorough understanding of the protein folding problem which in turn helps in understanding various aspects of biological functions. Characterizing the cooperative unfolding units and the intermediates along the folding funnel of a protein is a challenging task. We have investigated the native energy landscape of EhCaBP, a calcium sensor, belonging to the same EF-hand superfamily as that of calmodulin. EhCaBP is a two domain EF-hand protein consisting of two EF-hands in each domain and binding to 4 Ca²⁺. Native state hydrogen exchange (HX) was used to assess the folding features of the landscape and also to throw light on the structure-folding function paradigm of calcium sensor proteins. HX measurements under EX2 regime provided the thermodynamic information of the protein folding events under native conditions. HX studies revealed that the unfolding of EhCaBP is not a two-state process instead it proceeds through cooperative units. The C-terminal domain shows less denaturant dependence compared to that of its N-terminal domain suggesting that the former is dominated by local fluctuations. It is interesting to note that the N and the C-terminal domains of EhCaBP have distinct folding features. In fact, these observed differences may regulate the domain dependent target recognition of two-domain Ca²⁺ sensor proteins.

Differential Native State Ruggedness of the Two Ca²⁺-Binding Domains in a Ca²⁺ Sensor Protein.

Characterization of near native excited states of a protein provides insights into various biological functions such as co-operativity, protein-ligand and protein-protein interactions. In this project, we investigated the ruggedness of the native state of EhCaBP using nonlinear temperature dependence of backbone amide-proton chemical shifts. EhCaBP is a two-domain EF-hand calcium sensor protein consisting of two EF-hands in each domain and binds four calcium ions. It has been observed that ~30 percent of the residues in the protein access alternative conformations. Theoretical modeling suggested that these low energy excited states are within 2-3 kcal/mol from the native state. Further, it is interesting to note that the residues accessing alternative conformations are more dominated in the C-terminal domain compared to its N-terminal counterpart suggesting that the former is more rugged in its native state. These distinct characteristics of N- and C- terminal domains of a calcium sensor protein belonging to the superfamily of calmodulin would have implications for domain dependent Ca²⁺ signaling pathways.

NMR Studies on a Neuronal Calcium Sensor-1

Neuronal calcium sensor-1 (NCS-1), a Ca²⁺-binding protein plays an important role in the modulation of neurotransmitter release and phosphatidylinositol signaling pathway. It is known that the physiological activity of NCS-1 is governed by its myristoylation. Hence, we have studied the role of myristoylation in governing Ca²⁺ binding and Ca²⁺-induced conformational changes in NCS-1 as compared to the non-myristoylated protein, (in collaboration with Dr. Yogendra Sharma, CCMB, Hyderabad). ⁴⁵Ca binding and isothermal titration calorimetric (ITC) data show that myristoylation increases the degree of cooperativity and thus the myristoylated NCS-1 binds Ca²⁺ more stronger (with three Ca²⁺ binding sites) than the non-myristoylated one (with two Ca²⁺ binding sites). Both forms of protein showed different conformational features in far-UV CD when titrated with Ca²⁺. These results suggest that myristoylation influences the protein conformation and Ca²⁺-binding, which might be crucial for its physiological functions. Determining the solution structure of myristoylated and non-myristoylated NCS-1 by NMR will help dissect out common principles that govern structure-function-relationships of these proteins. This is of utmost importance, as the release of neurotransmitter and exocytosis, in general, are strongly regulated by the intracellular calcium concentration. As a first step to determine the 3D structure of both myristoylated and non-myristoylated NCS-1, we have already accomplished complete ¹H, ¹³C and ¹⁵N assignments in the holo form of myristoylated NCS-1 (BMRB Accession Number 6942) using state-of-the-art multi-dimensional NMR. This threw light upon various secondary structural elements present in the holo form of myristoylated NCS-1. We have studied the role of Mg²⁺ in Ca²⁺-binding to NCS-1.

How magnesium modulates calcium-binding, calcium-induced conformation and equilibrium unfolding transitions of neuronal calcium sensor-1?

Neuronal calcium sensor-1 (NCS-1) is a major modulator of Ca²⁺-signaling with a known role in neurotransmitter release. NCS-1 has one cryptic (EF1) and three functional EF-hand motifs (EF2, EF3 and EF4). However, it is not known which are the regulatory (Ca²⁺-specific) and structural (Ca²⁺ or Mg²⁺ binding) EF-hand motifs. For understanding the specialized functions of NCS-1, identification of ionic specificity of the sites is important. In this work, we have identified the specificity of Ca²⁺-binding and the role of Mg²⁺ in modulating Ca²⁺-binding to NCS-1. Ca²⁺ titration as monitored by ¹⁵N-¹H HSQC suggests that Ca²⁺ first binds to the EF2 and EF3 almost simultaneously followed by EF4. Our NMR data suggest that Mg²⁺ binds to the EF2 and EF3, thereby classifying them as structural sites, while EF4 is a Ca²⁺-specific or regulatory site. In the presence of Mg²⁺, Ca²⁺-binding induces unusual conformational rearrangements in the protein, and Ca²⁺ reverses the Mg²⁺ induced changes in Trp fluorescence and surface hydrophobicity dynamically. NCS-1 appears to be the only member of the calcium sensor family in which such a reversal of Trp fluorescence is documented upon Ca²⁺-binding in the presence of Mg²⁺. In a larger physiological perspective, the reduction of overall affinity of Ca²⁺ by Mg²⁺ from 90 nM to 440 nM would be advantageous to the molecule by facilitating the reversibility to the Ca²⁺-free state. Though the equilibrium unfolding transitions of apo and Mg²⁺-bound NCS-1 are similar, unfolding transitions of Ca²⁺-bound NCS-1 are different in the absence and presence of Mg²⁺. Thus, Mg²⁺ via structural EF-hand motifs plays a dynamic as well as structural role at the sub-domain level. This study demonstrates the importance of structural sites and underlines Mg²⁺ as a modulator of calcium homeostasis and active state behavior of NCS-1.

Please write me if you need any assistance or information (chary@tifr.res.in)