Bioinformatics Advance Access originally published online on December 5, 2006
Bioinformatics 2007 23(3):267-271; doi:10.1093/bioinformatics/btl617
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
In silico identification of putative metal binding motifs
Centre for Biotechnology, Anna University Chennai, 600 025, India
*To whom correspondence should be addressed.
 |
ABSTRACT |
|---|
 TOP ABSTRACT 1 INTRODUCTION2 METHODS3 RESULTS4 CONCLUSIONSREFERENCES |
|---|
Metal ion binding domains are found in proteins that mediate transport, buffering or detoxification of metal ions. In this study, we have performed an in silico analysis of metal binding proteins and have identified putative metal binding motifs for the ions of cadmium, cobalt, zinc, arsenic, mercury, magnesium, manganese, molybdenum and nickel. A pattern search against the UniProtKB/Swiss-Prot and UniProtKB/TrEMBL databases yielded true positives in each case showing the high-specificity of the motifs. Motifs were also validated against PDB structures and site directed mutagenesis studies.
Contact: pgautam{at}annauniv.edu; s_anishetty{at}yahoo.co.uk
|
1 INTRODUCTION |
|---|
|
TOPABSTRACT1 INTRODUCTION2 METHODS3 RESULTS4 CONCLUSIONSREFERENCES |
|---|
Bacteria have evolved several mechanisms to tolerate the uptake of heavy metal ions. These include efflux of metal ions outside the cell, accumulation and complexation of the metal ions inside the cell and reduction of heavy metal ions to a less toxic state (Nies, 1999). Metal resistant bacteria have been shown to have great potential in effective bioremediation of metal contaminated sites and bioleaching of ores.
Three different systems mediate efflux of divalent heavy metal cations from bacterial cells: resistance-nodulation cell division (RND) driven transenvelope exporters, cation diffusion facilitators (CDF) and P-type ATPases (Nies, 2003). The genes for the toxic metal resistance systems including metal ions Ag+, Cd2+, Co2+, Cu2+, Hg2+, Ni2+, Pb2+, Sb3+, Tl+, Zn2+ and oxyanions like AsO2–, AsO43–, TeO32– and CrO42– are usually plasmid or chromosome encoded (Silver, 1996).
Complexation of toxic metal ions by peptides or proteins also helps in the detoxification process. The tripeptide glutathione ‘GSH’ chelates and detoxifies Cd2+ in most organisms. In addition to this, GSH related peptides called phytochelatins chelate Cd2+ in plants and selected yeasts (Mehra and Winge, 1991; Vido et al., 2001). Reductases often detoxify metal ions by reducing them.
In recent times, there has been a growing interest in utilizing bacteria and fungi for inorganic material synthesis. Semiconductor nanocrystal formation has been reported in yeast and in filamentous fungi. Candida glabrata and Schizosachharomyces pombe produced CdS nanocrystals when cultured in the presence of cadmium salts (Dameron et al., 1989a,b). Nanoparticles of gold and silver were shown to be synthesized intracellularly in Verticullum fungal cells (Mukherjee et al., 2001). Extracellular formation of stable gold nanoparticles in water by the reduction of aqueous chloroaurate ions by the fungus Fusarium oxysporum has been reported (Ahmad et al., 2003). Nanocrystal formation by bacteria has also been reported. Silver nanoparticles of well defined size and shape have been precipitated in the periplasmic space of Pseudomonas stutzeri AG259 isolated from a silver mine (Klaus et al., 1999; Klaus-Joerger et al., 2001). More recently, intracellular CdS nanocrystals were synthesized on incubation of Escherichia coli with cadmium chloride and sodium sulphide (Sweeney et al., 2004).
Functional or structurally important regions in a protein family are well conserved across species. Several studies have addressed the issue of identification of functional residues in proteins and their importance in functional annotation (Livingston and Barton, 1996; Ouzounis et al., 1998; Pupko et al., 2002; George et al., 2005).
Multiple sequence alignment of representative proteins from different species helps in identifying such conserved regions. These can further be characterized as sequence motifs or patterns. Generating a functional motif involves identifying residues in a protein sequence that impart functional properties to the protein. Motifs when used to probe sequence databases help find and annotate members belonging to a particular family of proteins. With the completion of a number of genome sequencing projects, characterizing metal resistance proteins using a bioinformatic approach has become possible.
PROSITE (Falquet et al., 2002) is a database for motifs and patterns. PROSITE includes heavy-metal-associated domain signatures and profiles with PROSITE IDs PS50846 and PS01047 and an ATPASE_E1_E2 signature bearing the ID PS00154. These are broad spectrum motifs. In this study, specific putative metal binding motifs have been designed for the ions of the following metals cadmium, cobalt, zinc, mercury, arsenic, manganese, magnesium, molybdenum and nickel.
|
2 METHODS |
|---|
|
TOPABSTRACT1 INTRODUCTION2 METHODS3 RESULTS4 CONCLUSIONSREFERENCES |
|---|
Metal ion binding protein sequences for cadmium, cobalt, zinc, arsenic, mercury, molybdenum, manganese, magnesium and nickel were retrieved from the UniProtKB/Swiss-Prot (release 50.0) database (Bairoch et al., 2005) based on a gene name search. The number of hits obtained for each query was recorded after manual inspection.
Representatives of each metal binding protein set were subjected to a multiple sequence alignment using ClustalX (Thompson et al., 1997). Proteins that function in the uptake or detoxification of metal ions possess metal binding sites. Preferred ligands for soft and borderline metal ions like Cu2+, Zn2+ and Cd2+ are thiolates and amines (Frausto da Silva and Williams, 2001). Metal binding sites therefore contain cysteine or histidine residues. Asparatate and glutamate residues also co-ordinate metal binding. Most of the motifs have been designed around conserved cysteine, histidine, glutamate and asparatate residues.
The metal ion binding motifs were validated by performing a motif search against the UniProtKB/Swiss-Prot (release 50.0) and UniProtKB/TrEMBL (release 33.0) database using the PrositeScan tool (Gattiker et al., 2002) at PROSITE database. This tool also allows the user to give a user defined motif to be scanned against the specified databases. The hits obtained were recorded for each class of proteins. Representative structures were obtained from the Protein Data Bank (Berman et al., 2000) and the motifs were correlated with the metal binding residues. Where structural data were not available, information gathered from literature reports on mutant studies of the corresponding proteins was used to crosscheck the importance of the residues in the motifs.
|
3 RESULTS |
|---|
|
TOPABSTRACT1 INTRODUCTION2 METHODS3 RESULTS4 CONCLUSIONSREFERENCES |
|---|
Motifs were designed and validated for each protein family and are tabulated in Table 1. The metal, gene name, motifs designed, number of hits obtained from a motif search and the number of entries in UniProtKB Swiss-Prot and TrEMBL databases as obtained through a gene name search is given for each entry along with validation details. The PDB IDs of the representative structures used in the validation are presented in Table 1. Site directed mutagenesis studies which confirm the involvement of the residues in certain motifs, are also recorded. In most cases, the motifs were able to retrieve all instances of the corresponding protein. In cases where the number of hits obtained by a motif search is higher than the instances of the protein in UniProtKB, the additional hits were either related proteins or hypothetical proteins. For example, a gene name search for ArsA an arsenic efflux pump component retrieved 57 entries, whereas the motif designed for ArsA picked up 76 hits. A total of 10 out of these 76 hits were hypothetical proteins. In addition, there were a few UniProtKB entries where the gene name has not been assigned. However, our motif search picked up these proteins. Similarly, in some cases where the motifs missed some of the UniProtKB entries, some of the entries were fragments and it is plausible that the motif designed does not fall in the fragment.
|
The efflux of divalent heavy metal cations from bacterial cells is mediated through RND driven transenvelope exporters, CDF and P type ATPases (Nies, 2003). RND driven systems are protein complexes that span the complete cell wall of Gram negative bacteria. In the case of metal exporting RND driven efflux systems, the central pump of this protein complex is usually a member of the RND superfamily (Tseng et al., 1999). In addition to an outer membrane protein, a membrane fusion protein is also required to transport the substrate directly to the extracellular medium. On the other hand CDF proteins and P-type ATPases are single subunit transporters. They transport their substrates from the cytoplasm to the periplasm. CDF proteins are driven by chemiosmotic gradient formed by protons or potassium (Guffanti et al., 2002) whereas P-type ATPases hydrolyze ATP as their driving force (Lee et al., 2002).
The CzcCBA efflux pump mediates the efflux of Co2+, Zn2+ and Cd2+ (Rensing et al., 1997). It is a RND-driven transenvelope exporter with three components: CzcA a RND protein, CzcB a membrane fusion protein and CzcC an outer membrane protein. CzcD not only confers resistance to these metals but is also involved in the regulation of the CzcCBA pump. Motifs have been designed for CzcA and CzcC components of the CzcCBA pump. One of the motifs for CzcA has conserved aspartate and glutamate residues, which have been shown to be essential for metal transport through mutation studies (Goldberg et al., 1999). The motif ‘S[LK]AL[LI][SA]D[AS][GALV]H[MNS][LFA][ST]D’ for CzcD is a CDF signature and in addition to CzcD proteins also retrieves other CDF proteins. The histidine and aspartate residues in this motif are essential for the function of CzcD (Anton et al., 2004). The efflux of cadmium through the CzcCBA pump has been suggested as a plausible mechanism of cadmium sulphide nanocrystal formation in E.coli (Sweeney et al., 2004). Motifs have also been designed for a cobalt ion transporter CbiN and cobalamine biosynthesis protein CbiD.
Mercury resistance in bacteria is conferred through a plasmid borne mer operon. MerA encodes for mercuric reductase, which detoxifies Hg2+ by reducing it to volatile Hg. MerB, which encodes for an organomercurial lyase is found in plasmids of organisms showing broad spectrum resistance to organomercurials. MerT encodes for a transmembrane protein and is involved in mercury transport. MerP is a periplasmic mercury binding protein (Qian et al., 1998). MerC encodes for another mercury uptake protein (Hamlett et al., 1992; Sahlman et al., 1997). Cysteine, which is an important residue in mercury binding, is present in all the motifs that have been designed. MerA, B and P have been validated through PDB structures 1ZX9 (Ledwidge et al., 2005), 1S6L (Di Lello et al., 2004) and 1AFI (Steele et al., 1997), 2HQI (Qian et al., 1998), respectively. MerP of Bacillus species has a specific motif ‘CCPPSVV’. The motifs constructed for MerC are centered around conserved cysteines, which have been shown to be important for mercury uptake (Sahlman et al., 1997, 1999). Similarly, the cysteine pair in the MerT motif is essential for its function (Morby et al., 1995).
The arsenic resistance efflux system transports arsenite AsO33–, using a double protein ATPase system ArsA and ArsB acting in concert or through ArsB alone which functions as a chemiosmotic transporter. ArsC encodes for an enzyme that converts intracellular arsenate AsO43– to arsenite AsO33–, which is the substrate for the efflux system (Silver, 1996). The histidine residue in the motif designed for ArsA and the arginine and cysteine residues in the motifs for ArsC have been validated as being important for metal binding through PDB structures 1IHU (Zhou et al., 2001) and 1J9B (Martin et al., 2001), respectively. The motif for ArsB has been designed around a conserved aspartate.
A molybdenum ion binding motif designed for NifD, a MoFe nitrogenase component, contains a histidine, which is important for molybdenum ion binding (Strange et al., 2003). A molybdate ion MoO42– binding motif has been designed for ModE a molybdate dependent transcriptional regulator and is validated (Hall et al., 1999).
The motif ‘[DEN]LIP[LM]CHP[IVL]’ for molybdenum cofactor biosynthesis protein C (MoaC) has conserved cysteine and histidine residues, which are in the active site region as shown in PDB ID 1EKR (Wuebbens et al., 2000). Zinc is an essential trace element because of its role in catalytic and structural stability of many enzymes. However, at high-concentrations it can be toxic. In prokaryotes, homeostasis is maintained by regulating the uptake and the efflux of zinc ion. A conserved motif has been designed for ZnuB a membrane sector of the ZnuACB pump.
MntH and MntR proteins were taken as representative proteins for manganese. Both the motifs designed for MntH have been validated through reports on mutant studies (Kehres and Maguire, 2003; Haemig and Brooker, 2004). The glutamate residue in the motif for MnTR is important for metal binding as ascertained through PDB IDs 2F5C and 2F5D (Kleigman et al., 2006).
Magnesium chelatase subunit I (chlI) has been considered for magnesium. Cysteines in this protein are considered to be important for metal binding (Jensing et al., 2000) and one of the motifs has a conserved cysteine. The other two motifs for this protein have conserved histidine, aspartate and glutamate residues.
Nickel ion toxicity can be compared to that of cobalt. The best known nickel resistance is through a RND driven efflux system cnr (cobalt and nickel resistance) and ncc (nickel, cobalt and cadmium). Another high-affinity nickel transporter protein is NixA/NicT. Conserved motifs have been generated for the three classes of proteins. The motif for NixA has conserved histidine and aspartate residues. Site directed mutagenesis of these residues in NixA of Helicobacter pylori resulted in complete abolition of nickel uptake (Fulkerson et al., 1998).
ATP driven P-type heavy metal pumps represent a class of proteins that translocate toxic and essential metal ions across biological membranes. They are also called as The CPx type ATPases and form the ion translocation subclass of P-type ATPases. They contain a conserved CPC or CPX motif and varying number of C-Xaa-Xaa-C motifs in the N-terminal domain of the proteins (Solioz and Vulpe, 1996). They also possess a conserved aspartate, which is the phosphorylation site. Two signatures have been designed for the ion translocating P-type ATPases in this study. The motif ‘[IV]GDG[IV]NDAP[AT]LA’ is part of a hinge motif in P-type ATPases and site directed mutagenesis studies have shown that the conserved aspartate and proline residues are important for catalytic activity (Okkeri et al., 2002). The motif ‘CPC[AS]L’ is around conserved cysteine and proline residues. The conserved cysteines are important for metal resistance (Lowe et al., 2004). Both the motifs retrieve metal ion translocating P-type ATPases. Some of the motifs could not be validated due to lack of structural data or reports on mutant studies. However, since most of these motifs have residues considered to be important for metal binding, they could have a role in metal binding or structural integrity.
|
4 CONCLUSIONS |
|---|
|
TOPABSTRACT1 INTRODUCTION2 METHODS3 RESULTS4 CONCLUSIONSREFERENCES |
|---|
All the motifs that have been constructed in this study retrieved only the corresponding metal binding proteins from the UniProtKB indicating that they are all specific to the protein families taken into consideration. Specific motifs can be effectively used for functional annotation of proteins. Probing non-redundant databases with such motifs will help catalog organisms with a potential for metal tolerance. This knowledge can be used for developing efficient bioremediation strategies.
|
Acknowledgments |
|---|
The authors thank BTIS DBT Programme for computational facilities. P.G. and S.A. thank DST for funding through NSTI scheme, Grant No: SR/S5/NM-37/2005.
Conflict of Interest: none declared.
|
FOOTNOTES |
|---|
Associate Editor: Christos Ouzounis
Received on June 14, 2006; revised on November 14, 2006; accepted on November 29, 2006
|
REFERENCES |
|---|
|
TOPABSTRACT1 INTRODUCTION2 METHODS3 RESULTS4 CONCLUSIONSREFERENCES |
|---|
Ahmad, A., et al. (2003) Extracellular biosynthesis of silver nanoparticles using the fungus Fusarium oxysporum. Colloids Surf. B:Biointerfaces, 28, 313–318[CrossRef].
Anton, A., et al. (2004) Characteristics of zinc transport by two cation diffusion facilitators from Ralstonia metallidurans CH34 and Escherichia coli. J. Bacteriol, . 186, 7499–7507
Bairoch, A., et al. (2005) The Universal Protein Resource (UniProt). Nucleic Acids Res, . 33, D154–D159
Berman, H.M., et al. (2000) The Protein Data Bank. Nucleic Acids Res, . 28, 235–242
Dameron, C.T., et al. (1989a) Glutathione coated cadmium sulfide crystallites in Candida glabrata. J. Biol. Chem, . 264, 17355–17360
Dameron, C.T., et al. (1989b) Biosynthesis of cadmium-sulphide quantum semiconductor crystallites. Nature, 338, 596–597[CrossRef][ISI].
Di Lello, P., et al. (2004) NMR structural studies reveal a novel protein fold for MerB, the organomercurial lyase involved in the bacterial mercury resistance system. Biochemistry, 43, 8322–8332[CrossRef][Medline].
Falquet, L., et al. (2002) The PROSITE database, its status in 2002. Nucleic Acids Res, . 30, 235–238
Frausto da Silva, J.J.R. and Williams, R.J.P. The Biological Chemistry of the Elements: The Inorganic Chemistry of Life, (2001) 2nd edn , Oxford Oxford University Press.
Fulkerson, J.F., Jr, et al. (1998) Conserved residues and motifs in the Nix A protein of Helicobacter pylori are critical for high affinity transport of Nickel ions. J. Biol. Chem, . 273, 235–241
Gattiker, A., et al. (2002) ScanProsite: a reference implementation of a PROSITE scanning tool. Appl. Bioinformatics, 1, 107–108[Medline].
George, R.A., et al. (2005) Effective function annotation through catalytic residue conservation. Proc. Natl Acad. Sci. USA, 102, 12299–12304
Goldberg, M., et al. (1999) Energetics and topology of CzcA, a cation/proton antiporter of the resistance-nodulation-cell division protein family. J. Biol. Chem, . 274, 26065–26070
Guffanti, A., et al. (2002) An antiport mechanism for a member of the cation diffusion facilitator family: divalent cations efflux in exchange for K+ and H+. Mol. Microbiol, . 45, 145–153[CrossRef][ISI][Medline].
Haemig, H.A. and Brooker, R.J. (2004) Importance of conserved acidic residues in mntH, the Nramp homolog of Escherichia coli. J. Membr. Biol, . 201, 97–107[CrossRef][ISI][Medline].
Hall, D.R., et al. (1999) The high-resolution crystal structure of the molybdate-dependent transcriptional regulator (ModE) from Escherichia coli: a novel combination of domain folds. EMBO J, . 18, 1435–1446[CrossRef][ISI][Medline].
Hamlett, N.V., et al. (1992) Roles of the Tn21 merT, merP and merC gene products in mercury resistance and mercury binding. J. Bacteriol, . 174, 6377–6385
Jensen, P.E., et al. (2000) Modification of cysteine residues in the chlI and ChlH subunits of magnesium chelatase results in enzyme inactivation. Biochem. J, . 352, 435–441.
Kehres, D.G. and Maguire, M.E. (2003) Emerging themes in manganese transport, biochemistry and pathogenesis in bacteria. FEMS Microbiol. Rev, . 29, 263–290.
Klaus, T., et al. (1999) Silver based crystalline nanoparticles, microbially fabricated. Proc. Natl Acad. Sci. USA, 96, 13611–13614
Klaus-Joerger, T., et al. (2001) Bacteria as workers in the living factory: metal-accumulating bacteria and their potential for materials science. Trends Biotechnol, . 19, 15–20[CrossRef][ISI][Medline].
Kleigman, J.I., et al. (2006) A structural basis for the metal selective activation of the manganese transport regulator of Bacillus subtilis. Biochemistry, 45, 3493–3505[CrossRef][Medline].
Ledwidge, R., et al. (2005) NmerA, the metal binding domain of mercuric ion reductase, removes Hg(2+) from proteins, delivers it to the catalytic core, and protects cells under glutathione-depleted conditions. Biochemistry, 44, 11402–11416[CrossRef][Medline].
Lee, S.M., et al. (2002) Functional analysis of the Escherichia coli zinc transporter zitB. FEMS Microbiol. Lett, . 215, 273–278[CrossRef][ISI][Medline].
Livingstone, C.D. and Barton, G.J. (1996) Identification of functional residues and secondary structure from protein multiple sequence alignment. Meth. Enzymol, . 266, 497–512[ISI][Medline].
Lowe, J., et al. (2004) A mutational study in the transmembrane domain of Ccc2p, the yeast Cu(I)-ATPase, shows different roles for each Cys-Pro-Cys Cysteine. J. Biol. Chem, . 279, 25986–25994
Martin, P., et al. (2001) Insights into the structure, solvation, and mechanism of ArsC arsenate reductase, a novel arsenic detoxification enzyme. Structure, 9, 1071–1081[Medline].
Mehra, R.K. and Winge, D.R. (1991) Metal ion resistance in fungi: molecular mechanisms and their regulated expression. J. Cell. Biochem, . 45, 30–40[CrossRef][ISI][Medline].
Morby, A.P., et al. (1995) The role of cysteine residues in the transport of mercuric ions by the Tn501 MerT and MerP mercury-resistance proteins. Mol. Microbiol, . 17, 25–35[CrossRef][ISI][Medline].
Mukherjee, P., et al. (2001) Fungus-mediated synthesis of silver nanoparticles and their immobilization in the mycelial matrix: a novel biological approach to nanoparticle Synthesis. Nano Lett, . 1, 515–519[CrossRef].
Nies, D.H. (1999) Microbial heavy-metal resistance. Appl. Microbiol. Biotechnol, . 51, 730–750[CrossRef][ISI][Medline].
Nies, D.H. (2003) Efflux mediated heavy metal resistance in prokaryotes. FEMS Microbiol. Rev, . 27, 313–339[CrossRef][ISI][Medline].
Okkeri, J., et al. (2002) Introducing Wilson disease mutations into the zinc-transporting P-type ATPase of Escherichia coli. The mutation P634L in the ‘hinge motif’ (GDGXNDXP) perturbs the formation of the E2P state. Eur. J. Biochem, . 269, 1579–1586[ISI][Medline].
Ouzounis, C., et al. (1998) Are binding residues conserved? Pac. Symp. Biocomput, . 401–412.
Pupko, T., et al. (2002) Rate4Site: an algorithmic tool for the identification of functional regions in proteins by surface mapping of evolutionary determinants within their homologues. Bioinformatics, 18, S71–S77[Abstract].
Qian, H., et al. (1998) NMR solution structure of the oxidized form of MerP, a mercuric ion binding protein involved in bacterial mercuric ion resistance. Biochemistry, 37, 9316–9322[CrossRef][Medline].
Rensing, C., et al. (1997) New functions for the three subunits of the CzcCBA cation-proton-antiporter. J. Bacteriol, . 179, 6871–6879
Sahlman, L., et al. (1997) A mercuric ion uptake role for the integral inner membrane protein, MerC, involved in bacterial mercuric ion resistance. J. Biol. Chem, . 272, 29518–29526
Sahlman, L., et al. (1999) Roles of the four cysteine residues in the function of integral inner membrane Hg2+ -binding protein MerC. Biochem. Biophys. Res. Commun, . 255, 307–311[CrossRef][ISI][Medline].
Silver, S. (1996) Bacterial resistances to toxic metal ions—a review. Gene, 179, 9–19[CrossRef][ISI][Medline].
Solioz, M. and Vulpe, C. (1996) CPx-type ATPases: a class of P-type ATPases that pump heavy metals. Trends Biochem. Sci, . 21, 237–241[CrossRef][ISI][Medline].
Strange, R.W., et al. (2003) XAFS studies of nitrogenase: the MoFe and VFe proteins and the use of crystallographic coordinates in three-dimensional EXAFS data analysis. J. Synchrotron Radiat, . 10, 71–75[CrossRef][ISI][Medline].
Steele, R.A., et al. (1997) Structures of the reduced and mercury-bound forms of MerP, the periplasmic protein from the bacterial mercury detoxification system. Biochemistry, 36, 6885–6895[CrossRef][Medline].
Sweeney, R.Y., et al. (2004) Bacterial biosynthesis of cadmium sulfide nanocrystals. Chem. Biol, . 11, 1553–1559[CrossRef][ISI][Medline].
Thompson, J.D., et al. (1997) The Clustal_X windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Res, . 25, 4876–4882
Tseng, T.T., et al. (1999) The RND superfamily an ancient, ubiquitous and diverse family that includes human disease and development proteins. J. Mol. Microbiol. Biotechnol, . 1, 107–125[Medline].
Vido, K., et al. (2001) A proteome analysis of the cadmium response in Saccharomyces cerevisiae. J. Biol. Chem, . 276, 8469–8474
Wuebbens, M.M., et al. (2000) Insights into molybdenum cofactor deficiency provided by the crystal structure of the molybdenum cofactor biosynthesis protein MoaC. Structure, 8, 709–718[Medline].
Zhou, T., et al. (2001) Conformational changes in four regions of the Escherichia coli ArsA ATPase link ATP hydrolysis to ion translocation. J. Biol. Chem, . 276, 30414–30422
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||