RICERCA ITALIANA     

Supramolecular assemblies. The Dps (DNA-binding proteins from starved cells)- DNA and sorcin - calcium channels systems and their biological function

Università degli Studi di Roma "La Sapienza"

Abstract

The project concerns two systems differing in composition and function.

1. Dps proteins-DNA
The bacterial Dps proteins, expressed under a variety of stress conditions, protect DNA via two mechanisms, a physical one that entails sequestration of DNA in Dps-DNA complexes and a chemical one based on the Dps ferroxidase activity which inhibits formation of hydroxyl radicals from Fe(II). Escherichia coli Dps, the family prototype, interacts with DNA through the three lysine residues of the N-terminal region. In vitro experiments carried out within PRIN 2003 demonstrated that native Dps condenses plasmid DNA, i.e. forms large Dps-DNA complexes containing many Dps molecules and one or few plasmids, whereas mutants with one lysine in the N-terminus bind DNA without condensation. Moreover, unpublished data show that such lysine residues are partially methylated in Dps expressed during the stationary growth phase, but are unmethylated in the protein expressed during the exponential one. Methylation influences the mode of interaction with DNA, i.e. favors its condensation, and hence appears to act as a novel regulatory factor of potential importance in an evolutionary perspective. To define the mode of interaction between the two macromolecules at a resolution level never achieved before, the crystallization experiments which resulted in the production of X-ray quality albeit very fragile Dps-DNA co-crystals will continue; first their stability will be increased, thereafter methylated Dps-DNA complexes will be studied. The extent of methylation of the lysine residues at the N-terminus will be evaluated under different stress conditions (e.g. thermal shock, alkylating agents) to gain understanding of the functional ramifications of the reaction. The study of Thermosynechococcus elongatus Dps, the first Dps from thermophilic bacteria fits in this framework.

2. Sorcin-Ca2+ channels
Sorcin binds Ca2+ with high affinity and thereupon translocates from cytosol to cell membranes where it interacts with different target proteins in different tissues. The interaction with the ryanodine receptors (RyR) and the Na+-Ca2+ exchanger (NCX) in skeletal and cardiac muscle allows sorcin to regulate Ca2+ fluxes in such cells. Thus, sorcin inhibits RyR, whose activity increases the intracellular Ca2+ concentration by promoting Ca2+ release from the stores, and activates NCX which extrudes Ca2+ from cytoplasm. The research aims at characterizing the interaction between sorcin and the individual channels in molecular and physiological terms. The former are still largely unknown. Determination of the topology and of the kinetic and thermodynamic parameters of the binding reactions will establish how sorcin interacts in succession with the individual channels during the excitation-contraction cycle. The experiments will employ: wt sorcin, the N-terminal domain, the Ca2+-binding C-terminal one, whose structure was solved by the proponents, site-specific mutants of the EF3 hand which is endowed with the highest affinity for Ca2+ (e.g. E124A), of the D helix which is believed to transmit the information of Ca2+ binding from the EF3 hand to the rest of the molecule (e.g. W99G e W105G) and the F112 L mutant (in the region between the D helix and EF3) of interest due to its association with a form of familial hypertrophic cardiomyopathy; as Ca2+ channels, the ryanodine receptors RyR1 and RyR2 of skeletal and caridac muscle respectively, and NCX. The effect of sorcin, its domains and the various mutants on RyR will be assessed by measuring Ca2+ release in isolated cardiomyocytes, the effect on NCX will entail measurement of membrane currents in response to a voltage ramp both in the absence and in the presence of NiCl2 which inihibits NCX. Phage display will be used to identify the peptides forming the sorcin binding surfaces on RyR and NCX and surface plasmon resonance to determine the kinetic and thermodynamic parameters of the binding reactions. <<<

Principal Investigator

Emilia CHIANCONE Università degli Studi di ROMA "La Sapienza"

Research Objectives

The general aim of all studies on biological supramolecular structures, which involve mainly proteins and nucleic acids, is the understanding of the principles that determine the high specificity of the binding reaction, and of the energetics and dynamics of the process. The latter aspect is of importance since it allows the system to respond to the requirements of the living cell. The availability of a continuously increasing number of crystallographic structures of proteins and nucleic acids greatly facilitates the task. Indeed, the structural knowledge of the interacting surfaces allows one to plan and carry out focused experiments that permit the definition of the above mentioned parameters.
The specific aims of the project concern the structural and functional characterization of two different systems: the complexes between DNA and members of the Dps (DNA-binding Proteins from Starved cells) family and those between sorcin (Soluble Resistance-related Calcium-bINding protein) and Ca2+ channels. The proponent's group has already contributed to the analysis of their properties.
Dps proteins are expressed by bacteria under nutrient depletion in order to protect DNA from the attack of damaging factors (e.g. oxidants, alkylating agents or radicals). They represent one of the most important defence mechanisms during starvation and contribute to the enhancement of bacterial resistance during this and other stress conditions. The protective action of Dps proteins, all assembled as dodecameric shells of identical subunits, is achieved by means of a dual mechanism: i) the formation of Dps-DNA complexes that effectively sequester DNA and thereby provide a physical barrier to damaging agents and ii) the ferroxidase activity of the protein which inhibits the combination of Fe(II) with hydrogen peroxide that leads to the formation of the highly toxic hydroxyl radicals (5). In fact, all Dps proteins possess a very conserved ferroxidase center which catalyzes the oxidation of Fe(II) by hydrogen peroxide and the subsequent incorporation of Fe(III) in the protein internal cavity; the ferroxidase center was identified by the proponent's group in Listeria innocua Dps (3). In some bacteria the two mechanisms coexist, in others the Dps-DNA complexes can not be formed due to an N-terminus of reduced length and/or lacking the lysine residues that interact with DNA. In experiments conducted on N-terminus deletion mutants and wt Escherichia coli Dps, the proponent's group demonstrated recently that the number of N-terminal lysines determines the DNA condensation capacity of the protein, i.e. the formation of large complexes with plasmid DNA containing many Dps molecules and one or two plasmids (2). The aim is to arrive at a complete structural description of the complexes. To this end the crystallization experiments started in the framework of PRIN 2003 will continue in order to increase the stability of the X-ray quality Dps-DNA co-crystals already obtained.
A further aim concerns the biological role of Dps-DNA complex formation; it is related to the recent, unpublished observation that during the stationary growth phase of E. coli, when expression of Dps increases significantly, methylation of the N-terminal lysines takes place and thereby alters the mode of interaction with DNA. The regulative role of this methylation reaction will be addressed also in view of its evolutionary interest.
The sorcin-Ca2+ channels system has a different biological frame, namely the muscle excitation-contraction cycle. Sorcin, a PEF protein (Penta EF-hand Ca2+-binding proteins), has all the characteristics of the family, such as a two-domain organization of the polypeptide chain and the capacity to translocate reversibly in a Ca2+-dependent manner from cytoplasm to cell membranes where it interacts with different protein targets depending on the tissue. In skeletal and cardiac muscle such targets are membrane channels involved in excitation-contraction coupling, like the ryanodine receptors, the voltage-dependent L-type Ca2+ channels and the Na+-Ca2+ exchanger (NCX). RyR is involved in Ca2+ release from the stores into cytoplasm, whereas the physiological function of NCX consists in Ca2+ extrusion from the cytoplasm. Experiments on myocytes and cardiomyocytes indicate that sorcin regulates muscle fiber decontraction by inhibiting Ryr and activating NCX (17, 24). Given the fact that the molecular aspects of the individual interactions are still unknown, the project aims at their definition in order to gain understanding also of the molecular basis of pathological alterations of the heart due to sorcin mutations. The interacting surfaces will be identified and the Ca2+ concentration dependence of the kinetic and thermodynamic parameters of the reaction between sorcin and the individual channels determined. Both parameters are of importance since it may be surmised that during the muscle excitation-contraction-decontraction cycle sorcin binds to the various channels with a specific timing thereby allowing the regulation of Ca2+ fluxes in the cell.
In the proposed studies sorcin, the C-terminal Ca2+-binding domain, whose structure was determined by the proponent's group (12), the N-terminal one, and site-specific mutants of the EF3 hand endowed with the highest Ca2+ affinity (e.g. E124A) and of the D helix thought to be involved in transmitting the information of Ca2+ binding from EF3 to the rest of the molecule (e.g. W99G e W105G, available in the purified state). In addition, the F112L mutant will be characterized since this mutation localized in the region between the D helix and the Ca2+-binding loop of the EF3 hand is present in a familial form of hypertrophic cardiomyopathy (21). The characterization of the mutant, which has been expressed in E. coli and purified both as a whole protein and as the C-terminal domain, will contribute to the understanding of the structural basis of the pathological phenotype. Among the Ca2+ channels, the ryanodine receptors that are expressed principally in skeletal and cardiac muscle, RyR1 and RyR2 respectively, and NCX will be studied.
Binding of sorcin to Ryr involves the C-terminal domain (26); the fine topology of the interaction however has not been established even though the region around the D-helix is likely to be involved. The localized Ca2+ release events, named Ca2+ sparks, generated by the activity of Ryr will be measured in isolated cardiomyocytes perfused with sorcin, the C-terminal domain and the above-mentioned mutants. By comparing their effect on the frequency, amplitude and length of the Ca2+-sparks the specific region of the C-terminal domain involved in the interaction will be determined. In order to establish the sorcin region which interacts with NCX, in vitro experiments and isolated cardiomyocytes will be employed. In cardiomyocytes, the effect of sorcin, its domains and the mutants on membrane currents in response to a voltage ramp will be measured both in the absence and in the presence of NiCl2 which inhibits NCX.
The identification of the Ryr and NCX interacting surfaces is more complex due to the large dimensions of the cytoplasmic region of these membrane proteins. The phage display technique will be employed to select from phage libraries exposing random peptides of the target proteins those peptides that interact in a Ca2+-dependent manner with sorcin immobilized on different supports. The kinetic and thermodynamic parameters of the interaction established by the peptides thus identified with sorcin, as determined by means of surface plasmon resonance methods, will provide an overall picture of the network of interactions that is established in cardiomyocytes upon translocation of sorcin to membranes and thereby furnish a significant contribution to the understanding of the regulative functions exerted by the protein. <<<

Timescale

24 months

National and international background

The text of the B model is reported below since the proposed project comprises only one Research Unit.

The study of supramolecular structures is central to modern biology because the biological world is rich in such ordered molecular assemblies whose role is of fundamental physiological relevance. Examples that exemplify their nature, complexity and efficiency are oligomeric proteins and the extended actin and tubulin polymers of the cytoskeleton, the lipid bilayers of membranes, the multi-enzymatic complexes involved in cell respiration, and the protein-nucleic acids complexes of chromosomal structures. Supramolecular assemblies are stabilized by diverse weak interacting forces that contribute to the specific recognition of the different components and permit their dynamic assembly and disassembly in response to cellular cues; moreover, interactions with small molecules and ions may be of great structural and functional relevance.
The research project addresses the study of two different supramolecular systems, chosen on the basis of the knowledge acquired by the proponents thanks to grants from PRIN (1999, 2001 and 2003) and the National Research Council. The conceptual approach provides the unifying element, while the methodology differs in part; for this reason the two systems are presented separately.
The interest in the experimental systems chosen is proven by the numerous papers published in prestigious international journals, as apparent also in the cited bibliography.

1. Dps (DNA-binding Proteins from Starved cells) proteins - DNA

The Dps protein family is an important part of the complex defence mechanisms that become operative in bacteria under stress conditions in order to protect DNA, which is particularly exposed to the toxic action of oxidants and radicals due to the absence of nucleosomes. The first Dps protein was discovered in 1992 in Escherichia coli; it is induced under nutritional and oxidative stress and binds to DNA in a sequence-independent manner, hence the name. Under conditions of nutrient depletion, which are often accompanied by oxidative stress in the natural habitat of bacteria and which are realized in the laboratory during the stationary growth phase, DNA lesions can not be repaired through homologous recombination pathways, since stationary phase cells usually contain their chromosomes in one copy and the enzymes involved in the normal repair mechanims are degraded rather being synthesized. In E. coli, nutritional stress induces the synthesis and accumulation of Dps which becomes the principal chromatin component in the late stationary phase and protects DNA by segregating it in microcrystalline Dps-DNA complexes.
The interaction between Dps proteins and DNA does not take place by means of a known structural motif, but involves the N-terminal region of the polypeptide chain; such an involvement was hypothesized first on the basis of the X-ray crystal structure of E. coli Dps (1) and was later proven by the proponents (2). The E. coli Dps protein is a spherical dodecamer assembled from identical subunits that are characterized by a four-helix bundle tertiary structure. The highly flexible, disordered, positively charged N-terminal region, extends from the four helix bundle into solvent. In the dodecamer, endowed with 23 symmetry, the N-terminal regions are located in the vicinity of the three-fold symmetry channels that allow iron entry into the protein cavity. Their exposure to solvent permits interaction with the negatively charged DNA backbone. Noteworthy, the N-terminal regions have a regular spatial disposition, that is dictated by the symmetry of the dodecamer and is compatible with the formation of ordered Dps-DNA complexes such as those present in bacteria.
Studies on other members of the Dps family, which to date comprises over 50 proteins, have shown that many Dps proteins do not bing DNA due to the variability of the N-terminal sequence which can be very short and with no positively charged residues. In these proteins therefore DNA protection must be based on a different mechanism which was brought to light by the research activity on Listeria innocua Dps. L. innocua Dps is endowed with ferroxidase activity which is of utmost importance in protecting DNA against oxidative damage. As a matter of fact, all Dps proteins contain a highly conserved ferroxidase center and are able to catalyze the oxidation of Fe(II) utilizing hydrogen peroxide as oxidant and thereby reduce the toxic action of these two compounds which produces hydroxyl radicals (3-8).
The importance of Dps proteins as anti-oxidant agents is demonstrated also by "in vivo" studies, e.g. on Streptococcus mutans and Porphyromonas gingivalis, two microrganisms that lack catalase but are able to resist peroxide stress and whose survival upon exposure to reactive oxygen species is reduced by deletion of the dps gene (9, 10).
There are still a number of unresolved problems: for example the structure of the Dps-DNA complexes is not known nor whether such complexes have a more profound biological significance under specific bacterial growth conditions. These points will be addressed in the proposed research project.


2. Sorcin - Ca2+channels

Sorcin (Soluble Resistance-related Calcium-bInding proteiN) belongs to the PEF (Penta EF-hand) protein family and shares with all the family members the capacity to translocate in a reversible manner from cytosol to cellular membranes upon Ca2+-binding. The binding event triggers a conformational change that leads to the exposure of hydrophobic groups which in turn mediate the interaction with target proteins whose nature depends on the tissue where sorcin is expressed (11). It follows that like all other PEF proteins sorcin has a regulatory protein which to date has been elucidated only in part.
The sorcin polypeptide chain has a two domain organization: the N-terminal domain (1-32) is rich in glycines tyrosines and prolines, while the C-terminal one (33-198) contains the five EF hands, binds Ca2+ and is thus named SCBD (Sorcin Ca2+-Binding Domain). The SCBD crystallographic structure has been solved in the apo, Ca2+-free form (12). The mechanism of the Ca2+-induced sorcin activation is unusual since the physiological calcium binding sites, EF2 and EF3, are not coupled structurally as in all PEF proteins, but are separated by a long alfa-helix, the D helix (13). According to the proposed mechanism, the binding of Ca2+ to EF3, the highest affinity site, induces a rearrangement of the D helix which transmits the information to EF2 and therefrom to the rest of the molecule including the N-terminal domain. As a result, a reorientation of the two domains takes place which exposes hydrophobic regions to solvent and renders both domains accessible for interaction with target proteins. This mechanism is in accordance with sorcin's ability to interact simultaneously in vitro with two different proteins, namely annexin VII via the N-terminal domain and the ryanodine receptor (RyR) via the C-terminal one.
The interaction with annexin VII, a Ca2+-binding protein itself, takes place in vivo in adrenal medulla (14), but has an unknown physiological role because the function of annexin VII is likewise unknown. On the other hand thee interaction with RyR, the main Ca2+-channel of the sarcoplasmic reticulum, implies the involvement of sorcin in muscle contraction, a process where calcium has a well-known role as second messenger. The partecipation of sorcin in muscle contraction is confirmed by a number of data obtained in vivo. In isolated cells, Ca2+-bound sorcin appears to interact not only with RYR1 and RYR2 in skeletal muscle and cardiac cells respectively, but also with other Ca2+ channels involved in excitation-contraction coupling, such as the L-type Ca2+ channels, the Na+-Ca2+ exchanger (NCX), and the sarcoplasmic reticulum Ca2+-ATPase (SERCA2a) (15-18). A regulatory role of sorcin has been demonstrated for at least two such channels, Ryr2 and SERCA2a. Thus, sorcin activated by the increase of intracellular Ca2+ concentration, inhibits Ryr2 and activates SERCA2a thereby participating in the muscle fiber decontraction phase. However, many of the molecular aspects of the interactions established by sorcin with the different targets remain to be elucidated. <<<