Comparative structural analysis of cytokinin dehydrogenase enzymes of O.sativa, A. thaliana and Zea mays leading to predict the best enzyme

The cytokinin dehydrogenase enzymes of rice, wheat and maize were selected for this study. The enzymes of this crops show the same function i.e, leading to high grain production. The CKX1, CKX2, CKX3, CKX4, CKX5, CKX6, CKX7 and CKX8 enzymes of these three crops vary from each other on the basis of their physico-chemical analysis. A brief study has been done on all these enzymes showing the similarities, domains, secondary structures, homology models, backbone confirmations, best generated models and functions. From the study we got that Q4ADV8 of O.sativa has the best secondary structure with strong helix, turns and sheets along with highest molecular weight. A2XVN3 of O.sativa has the longest domain region, Q9FUJ1 of A.thaliana has highest numbers of total negatively charged residues (ASP+GLU) and Q9LY71 of A.thaliana has highest numbers of total positively charged residues (ARG+LYS). The Q9LTS3 of A.thaliana has highest VDW radius from geometric center. Q8LNV6 is the best model depending upon the Z-score calculated by ANOLEA. from the above study it is concluded that the enzymes of O.sativa andA.thaliana has more strong physico-chemical characteristics than in comparision to the zea mays.


Introduction
Due to the growth of population the demands for more grain production are increasing day by day. The main crops harvested in world are rice, maize and wheat. These are the main source of food. Human beings absorb many proteins from these crops. These crops are blessed with such miraculous proteins which help a lot in high grain production in order to fulfill the demands. Here we have made a brief study about the cytokinin dehydrogenase enzymes of rice, maize and wheat leading to high grain production. cytokinin dehydrogenase belongs tooxidoreductasefamily and catalyzes the chemical reaction. Cytokinin dehydrogenase is otherwise known as N6-dimethylallyladenine: (acceptor) oxidoreductase, 6-N-dimethylallyladenine: acceptor oxidoreductase, OsCKX2, CKX, and cytokinin oxidase/dehydrogenase. cytokinin dehydrogenase is helpfull in the degradation of cytokininsisopentenyladenine, zeatin, and their ribosides (chmülling et al., 2003).Gn1a gene of rice possessescytokinin oxidase/dehydrogenase (OsCKX2), an enzyme which degrades the phytohormonecytokinin (Ashikari et al., 2005). The Gn1a/OsCKX2 of Oryza sativa L, grain number 1a/Cytokinin oxidase 2 gene, encodes a cytokinin oxidase, which acts as a major quantitative trait locus contributing to grain number improvement in rice breeding practice (Lia et al., 2012). In wheat the CKX2 expression is activated by the IKU transcription factor WRKY10 directly and promotes endosperm growth (Li et al., 2013).In wheat the growth control of endosperm by CKX2 integrates genetic and epigenetic regulations. In angiosperms, cytokines are highly active in endosperm, IKU effectors coordinate environmental and physiological factors, resulting in modulation of seed size. Higher activity of the ckx3 ckx5 in wheat increase in seed yield, highlighting the relevance of sink strength as a yield factor. CKX3 and CKX5 regulate the activity of the reproductive meristems of Arabidopsis thaliana (Batrina et al., 2011). In this study we have considered the CKX1 of zea mays(Maize), CKX2 of Oryza Sativa Computational Molecular Biology 2014, Vol. 4, No. 14, 1-9 http://cmb.biopublisher.ca 2 (japonica), Arabidopsis thaliana (Mouse-ear cress), CKX3 of Oryza Sativa (japonica), Arabidopsis thaliana(Mouse-ear cress), CKX4 of Oryza Sativa (japonica), CKX5 of Arabidopsis thaliana(Mouse-ear cress), CKX6 of Arabidopsis thaliana (Mouse-ear cress) and CKX7 of Arabidopsis thaliana (Mouse-ear cress) and CKX8 of Oryza Sativa(indica).The main objective of the study is to find out their physic-chemical properties, similarities and differences among all the enzymes. However the function of all the eznzymes is same, but their properties are explained below.

Material and Methods
The cytokinin dehydrogenase enzymes of rice, wheat and maize were selected in this study due to its role in high grain production in crops.

Sequence retrieval
The cytokinin dehydrogenase enzymes of O.sativa (japonica), A.thaliana (mouse-ear cress) and zea mays (maize) were used for this study. Their amino acid sequences were retrived from uniprot (http://www. uniprot.org/), which is given in Table 1.

Q9LTS3
Catalyzes the oxidation of cytokinins, a family of N(6)-substituted adenine derivatives that are plant hormones, where the substituent is an isopentenyl group

Sequence alignment
The cytokinin dehydrogenase enzymes of all the plants were subjected for multiple sequence alignment using ClustalW2 (http://www.ebi.ac.uk/ Tools/msa/clustalw2/) analysis (Thompson et al., 1994). The alignment file was saved in clustal format and conserved residues were identified successfully. A phylogenetic tree was generated.
The conserved residues of all the enzymes are shown in Figure 1.

Physico-Chemical Characterization of all the enzymes
The amino acid sequences of all the enzymes were subjected to ProtParamtool (http://web.expasy.org/ protparam/) for further analysis, from where we got the physical and chemical properties of all the enzymes including molecular weight, theoretical PI, extinction coefficients, instability index, aliphatic index, grand average of hydrophobicity, total number of positively and negatively charged residues [9]. Detail information is shown in Table 2,  Table 3, Table 4.

Secondary structure prediction of enzymes
The CFFSP server (http://www.biogem.org/tool/ chou-fasman/) was used for secondary structure analysis where we submitted the amino acid sequences as an input and the result showed the details about helices, sheets and turns respectively (David et al., 1999). The detail information is shown in Table 5 and Figure 2-Figure 11.

Structural alignment
As the NMR crystallographic structures of these enzymes are not available at PDB, so the tertiary structures of all the enzymes of plants were designed by using phyre2 server whichgenerate reliable protein models when other widely used methods such as PSI-BLAST cannot. Phyre2 server (http://www.sbg. bio.ic.ac.uk/phyre2/html/page.cgi?id=index) generate the model using the principles and techniques of homology modeling. The target can be modelled with reasonable accuracy on a very distantly related sequence of template. The phyre2 server uses a profile-profile alignment algorithm based on each proteins position-specific scoring matrix. Phyre2 server include protein structure prediction, function prediction, domain prediction, domain boundary prediction, evolutionary classification of proteins, guiding site-directed mutagenesis and solving protein crystal structures by molecular replacement (Christie et al., 2012, Bilal et al., 2013, Singh et al., 2009. The homology models of enzymes are shown in Figure  12- Figure 21 and the detail information after visualization is shown in Table 6.

Model validation and optimization
The final tertiary structures of the enzymes were visualized usigPyMol visualization tool. Then by using PyMolvisualiser we got the number of atoms present in all the enzymes, formal charge sum, and partial charge sum, molecular and solvent accessible surface area. The backbone confirmation of all the enzymes were checked using Rampage server (mordred.bioc.cam.ac.uk/ ~rapper/rampage.php) from where we got the residues lying in allowed, favoured and outlier regions [13]. Then the models were subjected to ANOLEA-swiss model from where we got the Z-score, QMEAN score. The model with least Z-score is generally selected as the best model. The detail information is shown in Table 7.

Result and Discussion Sequence retrieval analysis
The amino acid sequences were successfully retrived from uniprot, which showed the following informations.

Multiple sequence alignment result
From the MSA analysis we got that there are many conserved residues among all the enzymes i.e, D, F, G, V, P, L, H, S, Q, W, Y, N, C, T, R, A, E, V, P, K, I. the similarities are denoted with *, dissimilarities with . and gaps with -.

Secondary structure prediction result of all enzymes
The secondary structures of all the enzymes were predicted using CFFSP server from where we could got the percentage of alpha, helix, turns. The helix is denoted with red colours, sheet is denoted with green colours, turn is denoted with blue colours. The secondary structures of all the enzymes are given from Figure 2- Figure 11.

Homology modeling results
The models which were generated by using phyre2 server were visualized by using PyMol visualization tool, which showed the following result.

Backbone confirmation result of enzymes
The final generated models of all the enzymes were submitted to RAMPAGE server from where we got the residues lying in allowed, favoured and outlier regions. The detail information is given below.

Discussions
From the above study we got the functions of all the enzymes, the most common residues among them, from the physico-chemical analysis we got that A2XVN3 of O.sativa has the most longest domain region. The best model which was selected on the basis of Z-score was Q8LNV6 ofO.sativa. All the enzymes has most common domain i.e, FAD binding PCMH type. Depending upon the back-bome confirmation of enzymes the best model was Q9T0N8