American Journal of Biochemistry and Molecular Biology

Year: 2011 | Volume: 1 | Issue: 4 | Page No.: 319-336
DOI: 10.3923/ajbmb.2011.319.336

In silico Characterization and Homology Modeling of Cyanobacterial Phosphoenolpyruvate Carboxylase Enzymes with Computational Tools and Bioinformatics Servers

Aubrey A. Smith and Manuela C. Plazas

Abstract: Phosphoenolpyruvate carboxylase (PEPC; EC4.1.1.31) catalyzes the irreversible β-carboxylation of Phosphoenolpyruvate (PEP) to yield Oxaloacetate (OAA) and inorganic Phosphate (Pi). PEPC contributes to photosynthetic and anaplerotic CO₂ fixation in higher plants and bacteria. The aim of this study was to determine the physicochemical properties of cyanobacterial PEPCs and to develop 3-dimensional models of selected enzymes. The biocomputational analyses were performed in silico using web-based software and servers. The alignment of cyanobacterial enzymes and secondary structure analysis revealed that there are conserved amino acid substitutions and polymorphisms between PEPCs from marine and fresh water organisms. Furthermore, some marine subgroups seem to possess unique amino acid stretches that may modulate various aspects of catalysis and regulation. Phosphoenolpyruvate carboxylase from Synechococcus PCC 7002, a marine organism, most closely resembles PEPC from fresh water organisms; for this reason, the enzyme was chosen for homology modeling alongside PEPCs from the fresh water strain Anabaena variabilis and the marine cyanobacterium Synechococcus RS 9917. Amino acids and domains that can distinguish between the fresh water and marine PEPCs were identified. Secondary structure analysis and homology modeling suggested that cyanobacterial PEPCs are primarily alpha helical with additional β-sheets flanking the characteristic central β-barrel. The physicochemical characteristics and the 3D models provide a framework for the purification and characterization of cyanobacterial PEPCs.

Keywords: synechococcus, proteomic tools, Phosphoenolpyruvate carboxylase, computational analysis and prochlorococcus

INTRODUCTION

Photosynthetic carbon fixation is the major means of carbon dioxide assimilation in marine ecosystems. These communities consist of photosynthetic prokaryotes and eukaryotes, namely phytoplanktons, chromophytes and cyanobacteria. These organisms use the Calvin-Benson-Bassham (CBB) pathway to metabolize CO₂ (Caldwell et al., 2007). The key enzyme of the CBB pathway is ribulose-1,5-bisphosphate carboxylase/oxygenase (RubisCO). RubisCO-based CO₂ fixation occurs in all cyanobacteria through C₃ intermediates. A significant amount of carbon flows into C₄ acids during photosynthesis. These compounds which normally account for less than 5% of CO₂ fixation rate in the dark, are found in significant quantities in cyanobacterial photosynthetic carbon fixation (Leegood, 2002). This indicates that a C₄ mechanism for inorganic carbon fixation is prevalent in cyanobacteria. Formation of the C₄ acid Oxaloacetate (OAA) occurs through the β-carboxylation of Phosphoenolpyruvate (PEP). OAA synthesis is catalyzed by three enzymes in plants and bacteria including PEP-carboxylase (PEPC) which is widespread in all higher plants and various types of bacteria (O’Leary et al., 2009). Cyanobacteria are prokaryotic organisms that possess the ability to perform oxygenic photosynthesis. Many are able to fix nitrogen and they are often used as fertilizers in (Soltani et al., 2007). Cyanobacteria are widely distributed and their habitats include oceans, soils, freshwater and extreme environments (Tomitani et al., 2006). Extremophilic strains have been found in various ecosystems (Karthikeyan and Gopalaswamy, 2009).

Cyanobacteria in general have the unique characteristic of possessing an incomplete Tricarboxylic Acid (TCA) cycle (Tripp et al., 2010). These organisms cannot derive metabolic energy from the Krebbs cycle because they lack α-ketoglutarate dehydrogenase and NADH oxidase. Because of this, PEPC has been assigned an anaplerotic role in cyanobacteria aside from its role as a key carbon fixation enzyme. Furthermore the C₄ compounds used to replenish the TCA cycle may also be required for the production of nitrogen storage molecules; such compounds may also be involved in organic stress response in plants and cyanobacteria (Khidir Ahmed, 2009). Gene expression and molecular characterization of PEPCs from C3 plants has not been extensively reported (Hammami et al., 2004). Several PEPCs have been purified from plants and bacteria. The best-described bacterial PEPC is that found in Escherichia coli; its three-dimensional structure has been determined at 2.8 A resolution (Kai et al., 1999) and its physical and chemical properties have been extensively investigated (Kwon et al., 2008).

Despite their importance in net carbon dioxide fixation cyanobacterial PEPC enzymes have not been highly studied. The pepc gene was initially cloned in the fresh water cyanobacterial strains Anacystis nidulans, Anabeana variabilis and Synechocystis PCC6301 (Chen et al., 2004; Izui et al., 2004; Knowles and Plaxton, 2003). While The E. coli enzyme displays cooperativity and is regulated by Fructose-1,6-bisphosphate (FBP) and Acetyl Coenzyme A (AcCoA), PEPC from the cyanobacterium Coccochloris peniocystis PEPC does not possess cooperativity and is inhibited by aspartate and TCA cycle intermediates (Chen et al., 2002). C. peniocystis PEPC resembles enzymes isolated from C₃ plants such as banana fruit PEPC. The latter is activated by hexose-monophosphates and malate, succinate, aspartate and glutamate activate it. The effect of those inhibitors varies with pH. PEPC from the cyanobacterium Anacystis nidulans is also a non-allosteric enzyme. The pepc gene from the oceanic cyanocaterium Synechcoccus PCC 7002 was also isolated and expressed (Smith et al., 2008).

The availability of internet based tools and a server provides an excellent opportunity to characterize the physicochemical properties of cyanobacterial PEPCs as well as their primary, secondary and three-dimensional structural properties. The purpose of this study was primarily to report the in silico analysis and characterization of cyanobacterial PEPCs.

MATERIALS AND METHODS

PEPC protein sequences: Phosphoenolpyruvate Carboxylase (PEPC) protein sequences were retrieved from the National Center for Biotechnology Information (NCBI) [www.ncbi.nlm.nih.gov]. The search yielded 146 bacterial sequences from which 27 cyanobacterial sequences were selected (Table 1).

E. coli and Zea mays sequence were also obtained from the NCBI database. The sequences were converted to FASTA format prior to analysis using the ReadSeQ sequence conversion server (Gilbert, 2003). The Database and sequence converters were last accessed in December 2010 from a computer terminal at Montgomery College, USA.

Sequence analysis
Sequence alignments: Multiple sequence alignments were performed with ClustalW (Larkin et al., 2007). The ClustalW alignment file was used to generate a shaded output with the BoxShade server (Fig. 2). Identical and similar amino acids were shaded black and grey respectively. Phylogenetic analysis of protein sequences was generated using the alignment obtained with ClustalW (Fig. 1). The ClustalW and Boxshade servers were last accessed in December 2010 from a computer terminal at Montgomery College, USA.

Structural analysis: The amino acid composition of the cyanobacterial PEPC sequences was computed using the PEPSTATS analysis tool (Rice et al., 2000). The physico-chemical parameters such as the molecular weight, isoelectric point (pI), extinction coefficient, half-life, aliphatic index, amino acid property, instability index and Grand Average Hydropathy (GRAVY) were calculated using the ProtParam tool of the Expasy proteomics server (Table 3). Secondary structure elements prediction was performed using the Network Protein Sequence Analysis (NPS@) server and the Secondary Structural Content Prediction (SSCP) server (Combet et al., 2000; Eisenhaber et al., 1996). The consensus secondary structure content and predicted disulfide patterns of each cyanobacterial PEPC are tabulated in Table 4. The presence of disulfide bridges was analyzed using the CYS-REC tool which predicts the most probable bonding patterns between available cysteine residues. The 3-D models of cyanobacterial PEPCs were constructed using the protein structure homology model building program SWISS-MODEL with energy minimization parameters (Arnold et al., 2006). The modeled tertiary structures were built on the basis of sequence homology with the high-resolution crystal structures of the E. coli and Zea mays enzymes. The Swiss PDB viewer (Guex and Peitsch, 1997) was used to visualize and refine the models and the standalone version of PyMOL was used to generate publishable images of the PEPC models. The modeled 3D structures were evaluated and validated with the WHAT IF and RAMPAGE programs (Lovell et al., 2003; Vriend, 1990). The aforementioned bioinformatics and modeling servers were accessed in between October 2009 and October 2010 from a computer terminal at Montgomery College, USA.

RESULTS AND DISCUSSION

The primary structure analysis of cyanobacterial phosphoenolpyruvate carboxylases indicates that the percentage of cysteine residues is less than 1% across species. The most abundant amino acid is leucine which makes up approximately 14% of the primary structures (Table 2). These trends were also observed in the E. coli and plant enzymes.

Despite the low cysteine content, the disulfide bridge prediction tool CYS-REC computed disulfide bonds in most cyanobacterial PEPCs (Table 4). There is no periodicity among leucine residues, so the presence of specific leucine-rich motifs or domains is unlikely. Leucine residues are found as dimers in cyanobacterial PEPCs; most significantly the sequences LLRGALL and LLEVLL are found in all cyanobacterial PEPCs but they are absent in the E. coli and plant enzymes. Furthermore, the sequence LLKRL is found in PEPC from fresh water cyanobacteria and Synechococcus PCC 7002 but it is absent in the bacterial and plant enzyme. Another leucine-rich stretch, LLLAK/QE is found in cyanobacterial PEPCs exclusively.

PEPC was slightly larger in fresh water organisms with an average molecular weight of approximately 118000 g mol^-1. Conversely, oceanic PEPCs had an average molecular weight of 114000 g mol^-1. The computed isoelectric point (pI) was below 7 indicating that the proteins will most likely precipitate in acidic buffers. The average extinction coefficient was 1.2 for fresh water enzymes and 1.0 for marine strains, the difference may be due to the higher percentage of tyrosine residues in fresh water enzymes. According to the ProtParam server, a protein whose Instability Index (Ii) is larger than 40 may be unstable therefore, the program predicted that all cyanobacterial PEPCs would be unstable in solution.

The aliphatic index which gives a measure of the relative volume occupied by alanine, valine, isoleucine and leucine, ranged from 92.62 to 97.67 for all cyanobacterial PEPCs. The relatively high Ai values indicate that cyanobacterial PEPCs maybe stable over a large range of temperatures. The Grand Average Hydropathy (GRAVY) values for PEPCs ranged from -0.287 to -0.432 indicating that the proteins will interact favorably with water (Table 3).

A consensus method was chosen in order to determine the secondary structure elements in each PEPC protein. The secondary structure consensus prediction program of the Protein Sequence Analysis server generated a secondary consensus where the most present predicted conformational state is reported for each amino acid. The methods used to generate the consensus were: The Double Prediction Method (DPM) (Deleage and Roux, 1987), Discrimination of protein Secondary structure Class (DSC) (King and Sternberg, 1996), Garnier Osguthorpe and Robson (GOR1) (Garnier et al., 1978), GOR3 (Biou et al., 1988), Hierarchical Neural Network (HNN) (Geourjon et al., 1999), PHDsec (Rost and Sander, 1993), the PredictProtein server (Rost et al., 2004), Predator (Frishman and Argos, 1996) and the Self-Optimized Prediction Method (SOPM) (Geourjon and Deleage, 1994).

The results of the consensus in Table 5 indicate that cyanobacterial PEPCs are largely alpha helical with less than 10% beta structures. This is in agreement with the X-ray crystal structure of plant and bacterial PEPCs (Matsumura et al., 2002).

A multiple sequence alignment of cyanobacterial, plant and E. coli phosphoenolpyruvate carboxylases was performed with ClustalW (Larkin et al., 2007). The alignment indicated that Synechococcus PCC 7002 PEPC shared the most identity with PEPCs from fresh water organisms and only 30% identity with E. coli and Zea mays PEPCs (Table 6). A phylogenetic tree of the alignment of cyanobacterial PEPCs was also generated; the cladogram showed that Synechococcus PCC 7002 PEPC is grouped with PEPC from fresh water cyanobacteria (Fig. 1). This is interesting since Synechococcus PCC 7002 is a marine organism. Furthermore, the alignment of cyanobacterial enzymes revealed that there are conserved amino acid substitutions between the marine and fresh water strains (Table 6).

Synechococcus PCC 7002 PEPC shows substitution patterns associated with the enzyme from the fresh water strains. Those single amino acid substitutions appear to distinguish PEPC from marine organisms and those from fresh water organisms (Smith et al., 2008). The multiple sequence alignment also revealed that there are stretches of amino acids that are exclusive to the marine and fresh water enzyme. In addition, stretches of amino acids that are unique to Prochlorococcus marinus str. MIT 9301, Prochlorococcus marinus str. AS9601, Prochlorococcus marinus str. MIT 9312, Prochlorococcus marinus str. MIT 9515 and Prochlorococcus marinus subsp. pastoris str. CCMP1986 have been identified (Fig. 2). In addition to Synechococcus PCC 7002 PEPC sharing most sequence homology with PEPC from fresh water organisms, the enzyme’s primary structure contains strictly conserved amino acids at positions unique to PEPCs of fresh water cyanobacteria. Those amino acids are listed alongside equally conserved residues found on marine PEPCs (Table 6). Aside from the listed amino acids, there are 20 other residues that share equivalent positions on Synechococcus PCC 7002 PEPC and PEPCs from fresh water cyanobacteria. Synechococcus PCC 7002 PEPC also shares homologous stretches of amino acids with fresh water enzymes rather than with the marine PEPCs (Fig. 2). The most homology between PEPCs is found at the C-terminal where there are only α-helices.

The homology models of phosphoenolpyruvate carboxylases from Anabaena variabilis, Synechococcus RS 9917 and Synechococcus PCC 7002 were modeled using the Zea mays enzyme as the template. Sequence homology scores of approximately 30% between the template and cyanobacterial sequences were sufficient to generate useful models. The stereochemical and energetic properties of the models were evaluated with the WHAT IF and RAMPAGE servers. All three models were deemed acceptable by the structure validation criteria within the WHAT IF server. According to the Ramachandran plots, 87 to 88% of residues were in the most favored regions, 8.1 to 8.7% of residues were in the allowed regions while 3.7 to 4.0% of residues were in generously allowed regions. These results indicate that the models were geometrically viable. The cyanobacterial PEPCs contain the characteristic β-barrel seen in all PEPCs (Fig. 3). Residues that are important in bicarbonate binding as well as those involved in catalysis are found near the C-terminal of the β-barrel (Kai et al., 1999). The modeled Synechococcus RS 9917 PEPC only contains this β-barrel as its lone β structure. On the other hand, Anabaena variabilis and Synechococcus PCC 7002 PEPCs contain one and two additional β-sheets, respectively (Fig. 3). One of the β-sheets of Synechococcus PCC 7002 PEPC lies just below the N-terminal side of the β-barrel (Fig. 4).

Residues that were found to be essential for catalysis and regulation in the E. coli and Zea mays enzymes are conserved in fresh water and marine cyanobacterial PEPCs. The enzyme’s active site has been found to be made of H138, R396, K546, H579, R581, R587 and R699 (E. coli numbering). A loop region, GRGGSIGRGG (GRGGSVGRGG in Synechococcus PCC 7002) is involved in catalysis and binding to aspartate, an allosteric inhibitor (Matsumura et al., 2002). Another loop involved in catalysis and bicarbonate binding, KRRP(G/T)GG, is found in E. coli and plant isoforms but it is absent in the cyanobacterial enzyme. Furthermore, E433 and R438 (E. coli numbering) which are involved in intersubunit contact to form the tetramer are conserved in all PEPCs.

All PEPCs contain the aspartate binding site homology which is composed of three domains: EM(T/V)(L/F)(S/A)K, LRN(G/I)(T/Y) and MRNTG. There may be other residues and domains involved in regulation by aspartate as Anabaena variabilis is not subject to control by aspartate (Izui et al., 2004). The following sequences were found in all cyanobacterial PEPCs but they were absent in E. coli and plant PEPCs: ¹⁵⁰LNVPPX(Q/K)X(E/Q)XL¹⁶⁰, ²⁴⁵VDYALHYFQEVLF²⁵⁷, ³⁰⁸WXTACYQR³¹⁵, ³³⁶S(M/L)(H/Q)W(S/C)XVXXXLLESLE³⁵¹ (Synechococcus PCC 7002 numbering). In spite of the low sequence homology between cyanobacterial PEPCs and the allosterically regulated E. coli and Maize enzymes, the previously reported domains and residues involved in catalysis are shared by all PEPC proteins analyzed thus far. Unique structural features of PEPC may afford cyanobacteria with a CO₂ concentrating mechanism that is similar to what is found in plants (Sikolia et al., 2009).

CONCLUSION

In this study cyanobacterial phosphoenolpyruvate carboxylases were selected and characterized from a physicochemical perspective. For these enzymes, molecular weight, theoretical isoelectric point, molar extinction coefficient, aliphatic index, instability index, number of negative residues, Grand Average Hydropathy (GRAVY) and disulfide bond proclivity were computed. Physicochemical parameters provided useful data for the purification of cyanobacterial PEPCs. The primary structure of cyanobacterial PEPCs was further analyzed and amino acids as well as domains were identified that can distinguish between the fresh water and marine enzyme. Secondary structure analysis revealed that the proteins are largely alpha helical; this was supported by the homology modeling of PEPCs from Synechococcus RS 9917, Anabaena variabilis and Synechococcus PCC 7002. The latter is a marine organism but its PEPC is most similar to PEPCs from fresh water cyanobacteria in all aspects. The 3D model of Synechococcus PCC 7002 PEPC suggests that the molecule contain two β-sheets in addition to the central β-barrel. Those β-sheets may play a role in regulation and catalysis. While mutagenesis will be crucial in determining the function of the various residues and structures identified in this study, the 3D models can be used for functional analysis until crystal structures become available for cyanobacterial PEPCs.

REFERENCES