Evolutionary and functional analysis of fructose bisphosphate aldolase of plant parasitic nematodes

The essential and ubiquitous enzyme fructose bisphosphate aldolase (FBPA) has been a good target for controlling the various types of infections caused by pathogens and parasites. The parasitic infections of nematodes are the major concern of scientific community, leading to biochemical characterization of this enzyme. In this work we have developed a small dataset of all types of FBPA sequences collected from publically available databases (EMBL, NCBI and Uni-Port). The Phylogenetic study shows that evolutionary relationships among sequences of FBPA are clustered into three main groups. FBPA sequences of Globodera rostochiensis (FBPA_GR) and Heterodera glycines (FBPA_HG) are placed in group II, sharing the similar evolutionary relationship. The catalytic mechanism of these enzymes depends upon which class of aldolase, it belongs. The class of enzyme has been confirmed on the basis of sequences and structural similarity with template structure of class I FBPA. To confirm catalytic mechanism of above said model structures, the known substrate fructose-1, 6-bisphosphate (FBP) and competitive inhibitor Mannitol-1, 6 bisphosphate (MBP) were docked at known catalytic site of enzyme of interest. The comparative docking analysis shows that enzyme-substrate complex is forming similar Schiff base intermediate and conducts C3–C4 bond cleavage by forming Hydrogen bonding with reaction catalyzing Glu-191, reactive Lys-150, and Schiff base forming Lys-233. On the other hand enzymeinhibitor noncovalent complex is forming cabinolamine precursor and the proton transfer by the formation of hydrogen bond between MBP O2 with Glu191 enabling stabilization of cabinolamine transition state, which confirms the similar inhibition mechanism. Thus we conclude that Plant Parasitic Nematodes (PPNs) have evolutionary and functional relationship with the class I aldolase enzyme. Hence, FBPA can be targeted to control plant parasitic nematodes.

catalyzes the segmentation of structurally related sugar phosphates including fructose 1-phosphate (Fru 1-P) an intermediate of fructose metabolism. FBPA has standard classification number EC 4.1.2.13 and classified into two classes: Class-I and Class-II. With little exception in both class, aldolase are characterized more in vertebrates and lesser in invertebrates [4,5]. Tissue specific class I aldolase are found in animals, plants, green algae and other higher organism [6]. Three unique forms of class I aldolase have been determined in various tissues of vertebral species, including human. These are aldolase A (found in skeletal muscle and red blood cells), aldolase B (found in small intestine, liver and kidney), and aldolase C (found in smooth muscle and neuronal tissues) [7]. These forms are differentiable on the basis of immunological and kinetic properties [8]. The class-II aldolase is commonly present in lower organism e.g. bacteria, yeasts, fungi etc. Class-II is divided into group A and B depending upon insertion or deletion in amino acid sequences. Group A, enzymes play major function in glycolysis and gluconeogenesis, while group B enzymes are more heterogeneous and has diverse metabolic roles and substrate specificities [9].
The catalytic mechanism of class I and class II aldolase have been extensively studied and it is different in both classes. The class I enzymes use reactive lysine residue in the active site to stabilize a reaction intermediate via Schiff-base formation [10,11]. While the class-II enzymes polarize the substrate carbonyl by using divalent metal ion generally zinc. Although tertiary structures of both classes form the (β/α)8 barrel fold, also known as the TIM barrel fold, which shares the same overall fold but do not share any significant sequence homology or common catalytic residues as well as distinct location of their active sites [12]. In present work, we have found out the evolutionary relationship between all available characterized aldolase sequences from nematodes till date. In addition, belonging class and their catalytic mechanism of PPNs aldolase sequences have been dug out using comparative sequence analysis and structural biology approaches.

Methodology: Collection of FBPA protein sequences
The EMBL, NCBI and Uni-port protein databases were searched with individual key word like aldolase, Fructose-bisphosphate aldolase and Fructose 1, 6-bisphosphate aldolase of nematodes, browsed protein sequences were screened and downloaded. All incomplete sequences were excluded from dataset. Two or more complete sequences which were having 100% sequence identity as determined by the EMBOSS pair wise alignment tool, were removed from dataset [13]. For each aldolase sequence the locus number, name of the protein, experimentally determined functions and associated references were collected. In our dataset each sequence was allocated a specific name which has a reference code followed by scientific name of the nematodes.

Sequence and Phylogenetic analysis
Multiple sequence alignment (MSA) of dataset sequences were performed by online ClustalW2 Tool [14]. For initial pair wise alignment, Gonnet protein weight matrix (GPWM) with gap opening penalty 10 and a gap extension penalty 0.1 have been used. For multiple alignments, gap opening penalty 10, gap extension penalty 0.1, GPWM and 5 tree iteration (by Neighbor-Joining method) have been used. To assist comparison between PPNs (Globodera rostochiensis (FBPA_GR) and Heterodera glycines (FBPA_HG)) and other nematodes sequences were included in the alignment. It helped us to find out the percentage similarity of FBPA sequences of PPNs with other dataset sequences. The alignment file viewed and Phylogenetic tree was generated using percentage sequence identity with neighbour-joining algorithm in the Jalview-multiple sequence editor Tool [15].

Tertiary structure prediction
The functions of both class enzymes mainly depends upon formation of the TIM barrel fold and present residues in active site. To find out functions of dataset aldolase protein sequences required to generate suitable molecular model. The dataset protein sequences  [17]. The best template for each sequence was selected on the basis of similarity, percentage of identity, expectation value, bit scores and query coverage area. The best template structures for each sequence with their validation statistics depicted in Table 1.
The Chain A of FBPA from Rabbit Muscle (PDB ID: 1ZAH) was common template for sequences except Nematocida parisii sequences [18]. The 1ZAH structure was generated by X-ray diffraction study with 1.80Å resolution. The R factor of the structure was 0.167 and R-free value was 0.205. We have generated molecular models for plant-parasitic nematodes only as they are our prime concern. The models were generated using Modeller 9v10 [19] and validated using online server Structural analysis and verification server (SAVS) [20-23]. Secondary structural investigation and conformation analysis of modeled structures were performed by ProFunc, an online server of PDBsum [24]. The comparative sequence and structural analysis of active site and formation of TIM barrel fold for modeled structures with template structure have been analyzed by ClustalW2 tool and USFChimera.

Simulation and validation of models
The sequences and structural comparison between generated models of FBA_HG shows 99.9% sequence similarity with similar active site residues and sharing same homology of 3D Structure. To find out stable active sites and role of TIM barrel fold in catalytic mechanism, we performed simulation of model FBPA_HG structure using GROMACS 4.5.3 package [25,26]. To understand mechanism with exact comparison, we also performed molecular dynamics (MD) for template structure 1ZAH_A. Selected both structure were subjected for energy minimization using OPLS-AA/L force field [27]. In the subsequent steps the structure of FBA_HG and 1ZAH_A were embedded in a cubic box containing SPC216 water molecules [28]. Normal charge states of ionizable groups at pH 7 for FBPA_HG and 1ZAH_A have been neutralized by adding respective ions in the system. Energy minimization was performed after this ion treatment. The next step of the protocol was to maintain the equilibrium of the system which was performed in two phases. The first phase include the NVT ensemble, a short 100 picoseconds (ps) position-restrained MD simulation at 300K was carried out using a Berendsen thermostat to ensure the proper stabilization of the temperature. The second phase include NPT ensemble for 100ps position-restrained MD simulation at 300K and 1 bar was carried out using a Parrinello-Rahman barostat pressure coupling to stabilize the system with respect to pressure and density [29]. Finally unrestrained 10 nanoseconds (ns) MD simulation was initiated on the NPT ensemble for both structures. The quality checks on the MD simulations were performed by GROMACS applications. The numerical graphs and interpretation of data were performed using Xmgrace software. The structures which were qualifying the all validation parameters with good scores were further subjected for docking analysis.

Results: Sequence collection and Phylogenetic analysis
The collection of available aldolase sequences of nematodes from the protein databases was an initial step towards developing a comprehensive small dataset. BLAST was used to search other amino acid sequences of aldolase in order to obtain all available sequences from the NCBI, EMBL and Swiss-port. Searching results shows all aldolase sequences from nematodes as well as nematodes keyword associated organism's aldolase sequences. Since our main focus was on aldolase sequences of nematodes, were downloaded and saved in a form of small dataset. The partial and fragmented sequences have been removed manually. The identical sequences of same nematodes Spp. removed from the dataset after crosschecking in EMBOSS [13]. The screened dataset contained twenty-four unique FBPA Protein sequences given in Table 1. Phylogenetic analysis was performed to find out the Evolutionary relationships among dataset sequences and generated tree among each sequence shown in (Figure 1) [15].
The tree is classified into three major groups or cluster (namely I, II and III). The first, second and third cluster contains ten, eight and five sequences respectively. FBPA sequences of the PPNs clustered in group-II as shown in (Figure 1). This group contains FBPA proteins of Heterodera glycines (two), Globodera rostochiensis (one), Ascaris suum (one), Bursaphelenchus xylophilus (one), Caenorhabditis elegans (two) and Caenorhabditis brenneri (one). Phylogenetic tree inferring FBPA sequences of PPNs having similar evolutionary relationship with group-II and slightly differ with group-I and III.

Structure models and active site identification
The structural alignment result for all FBPA sequences using DELTA-BLAST against PDB data base shown in shows better sequence-to-structure agreement in comparison to initial proteins shown in Table 2 [23]. The overall quality Gfactor score for 1ZAH_Am and FBPA_HGm were -0.30, and -0.17 respectively, indicating good quality models. A detailed secondary structural investigation of FBPA_HG with PDBsum [24] shows that monomer unit of structure folds into 13 Alfahelices, 10 Strands, 19 Beta-turns and 2 Gama-turns depicted in ( Figure 3B). The tertiary structure FBPA_HG shows close resemblance to crystal 1ZAH_A and having 0.507 Å RMSD shown in (Figure 3A). Low RMSD and validation statistics reflects the high structural conservation of model structure through evolution.

Molecular dynamics simulation analysis of structures
MD simulations were performed to get stable structure of 1ZAH and FBA_HG. The main-chain root mean square deviations (RMSD) were calculated for both structures as a function of time. The resulting RMSD profiles are shown in ( Figure 4A).  Figure 4B). Both the proteins display a similar fluctuation pattern except for C terminal and N terminal regions. All catalytic site residue atoms are sharing the similar fluctuation pattern for both structure depicted in ( Figure 4B). However, the minimum fluctuation in substrate binding site in both proteins, leading to make a good observation of catalytic mechanism during docking studies. The solvent accessible surface area (SASA) for both structures is accessible to a solvent and it can be related to the hydrophobic core. The results indicate that the hydrophobic cores for both structures are in the range of ~85.0-95.0nm 2 and the SASA of FBPA_HG is much higher than 1ZAH_A shown in ( Figure 4C).

Docking analysis
The FBPA_HG enzyme adopt the same catalytic mechanism as of class I aldolase enzyme. A covalent catalysis entailing a Schiff base formed between a lysine residues of the enzyme and ketose substrate. According to mechanism the ketose (2)  Comparative docking results of MBP inhibitor at the catalytic site of both enzymes are depicted in Table 3. Both docked conformer of 1ZAH_Am and FBPA_HGm shown in (Figure 5B & D) are forming nine Hydrogen bonds and having similar binding energy. The MBP inhibitor is forming the Hydrogen bonds with Glu-191 and Glu-187 of FBPA_HGm and 1ZAH_Am respectively. These results also support to conclude that FBPA_GR is following the similar substrate recognitions and enzyme inhibition mechanism as class I aldolase.

Conclusion:
This study was carried out to find the evolutionary and functional relationship of FBPA proteins of plant parasitic nematodes. The respective protein sequences are collected and after sequential refinement, a small dataset is built. The Phylogenetic analysis of FBPA protein sequences are clustered into three major groups according to their sequence similarity. The structural alignment of theses sequences against PDB database confirms that collected sequences belongs to class I types of aldolase. To get insight of the catalytic mechanism of Heterodera glycines and Globodera rostochiensis, molecular model were generated and validated. In subsequent step molecular dynamics simulation was performed to obtain stable structure of 1ZAH and FBPA_HG. The catalytic mechanism of FBPA_HG has been discovered on the ground by docking of substrate FBP and a competitive inhibitor MBP at the active site of both enzymes. The results confirmed that FBAP_HG and FBAP_GR are following the similar enzyme-substrate and enzymeinhibitor reaction geometry and reaction intermediate same as class I 1ZAH aldolase. This work may be helpful to experimental biologist in controlling the parasitic infections by inhibition of aldolase.