|In Silico Biology 8, 0034 (2008); ©2008, Bioinformation Systems e.V.|
1 Western Australian Biomedical Research Institute
2 School of Biomedical Sciences
3 School of Pharmacy, Curtin University of Technology, GPO Box U1987 Perth, WA 6845, Australia
* Corresponding author
Phone: +61-8-92661017; Fax: +61-8-92662342
Edited by H. Michael; received May 10, 2008; revised August 13, 2008; accepted August 14, 2008; published September 19, 2008
The incidence of human infections by the fungal pathogen Candida species has been increasing in recent years. Enolase is an essential protein in fungal metabolism. Sequence data is available for human and a number of medically important fungal species. An understanding of the structural and functional features of fungal enolases may provide the structural basis for their use as a target for the development of new anti-fungal drugs. We have obtained the sequence of the enolase of Candida krusei (C. krusei), as it is a significant medically important fungal pathogen. We have then used multiple sequence alignments with various enolase isoforms in order to identify C. krusei specific amino acid residues. The phylogenetic tree of enolases shows that the C. krusei enolase assembles on the tree with the fungal genes. Importantly, C. krusei lacks four amino acids in the active site compared to human enolase, as revealed by multiple sequence alignments. These differences in the substrate binding site may be exploited for the design of new anti-fungal drugs to selectively block this enzyme. The lack of the important amino acids in the active site also indicates that C. krusei enolase might have evolved as a member of a mechanistically diverse enolase superfamily catalying somewhat different reactions.
Keywords: enolase, Candida, phylogenetic tree, plasminogen, anti-fungal activities
Invasive or systemic fungal infections are now a major cause of morbidity and mortality in hospital patients. Developments such as intensive cancer chemotherapies, organ transplantation and the use of broad spectrum antibiotics have contributed to the rising levels of fungal infections . Candida spp. are now the fourth most commonly isolated pathogen in bloodstream infections . Other fungal species such as Aspergillus and Cryptococcus are also frequently associated with particular medical conditions such as bone marrow transplantation and HIV/AIDS .
There are a number of classes of anti-fungal drugs currently available to treat fungal infections. These classes include the azoles, polyene, allymines and the echinocandins . Most of these drugs target fungal cell wall or cell membrane synthesis, and interfere with the growth of the organism, rather than actually kill it. Current anti-fungal therapy is hampered as there is no single broad spectrum, rapidly fungicidal drug available which has low toxicity and other desirable pharmacokinetic properties. The azole drugs, especially fluconazole are the most commonly used anti-fungal drugs . While the azole drugs have good pharmacokinetic properties and low toxicity, they are fungistatic, meaning that they do not kill the pathogen but rather they inhibit its growth. Resistance to anti-fungal drugs is also a significant problem. For example, C. krusei is considered resistant to the azole class of anti-fungals and some other Candida species particularly Candida lusitaniae, exhibit resistance to the polyene drug, amphotericin B . Due to the limitations of these currently available drugs, the mortality rate due to fungal infections remains very high .
In order to provide effective therapy, any new anti-fungal drug needs to display broad spectrum fungicidal properties as well as low toxicity and good pharmacokinetic properties. The development of a fungicidal drug requires the identification of a drug target that is an essential fungal protein. Enolase is a protein that plays a vital role in glycolysis and therefore energy production within the cell . Targeting the fungal enolase selectively would lead to the rapid death of fungal cells with little or no toxicity to the human host. Prior to seeking to obtain this selective inhibition, it is necessary to determine if there are any amino acid residue differences in and around the active site of human and fungal enolases. These amino acids could then be utilised to design a ligand to selectively bind and inhibit the function of fungal enolase, thus forming the basis for the creation of a new anti-fungal drug class. In this research, we report the sequencing of C. krusei enolase in order to determine any structural differences with respect to human enolase that may be exploited for anti-fungal design.
Isolation of C. krusei DNA
C. krusei was grown in 50 ml of YEPD broth (1g yeast extract, 0.5 g peptone and 6 ml of 20% [wt/vol.] glucose solution). Aliquots of the culture were inoculated on CHOMagar plates and ID32 test strips to confirm the yeast identity. Upon confirmation, a fresh culture was grown for DNA extraction using the yeast genomic DNA preparation protocol described by Hoskins and Hahn .
Generation of C. krusei PCR primers
Degenerate PCR primers EnoFor1 and EnoRev2 were designed by the alignment and comparison of previously published enolase sequences from yeast species Saccharomyes cerevisiae (S. cerevisiae) and Candida albicans (C. albicans). Both sequences were obtained through GenBank (accession numbers JO_1323 and M93712 respectively). The primer sequences were derived from conserved sequences near the start and end of the enolase gene coding sequence. A BamH1 restriction enzyme recognition sequence was added at the 5' end of each primer to facilitate the subsequent cloning. Primers were synthesised by GeneWorks Ltd (Adelaide, South Australia) with the following sequences: EnoFor1- GCT CCG GAT CCC GTC TAC GAC TCY MGW GGT AAC CC and EnoRev2- GCT CCG GAT CCG GAA RTY YTB RCC RGC GTA GAY RGC, whereby R = A or G, Y = C or T, M = A or C, and W = A or T.
Amplification of the C. krusei enolase gene
The PCR reaction mix containing 100 μl PCR mastermix (0.2 mmol of each deoxynucleotide, buffer and 1.5 mmol of MgCl2), 1 μl (of 1 μg/μl) of each forward (EnoFor1) and reverse (EnoRev2) primer, and 1 μl (5 units) of Taq DNA polymerase was divided into ten 9 μl aliquots. One microlitre of C. krusei genomic DNA was added as either a neat, 1 in 5 or 1 in 10 dilution to make the final reaction mixture 10 μl. The cycling parameters were 95°C for 6 min: 30 cycles of denaturation for 30 sec at 95°C, annealing for 30 sec at 55°C, an extension for 1 min at 72°C, with a final extension of 5 mins.
Cloning of C. krusei enolase
The resulting PCR product was ligated into the PGEM-T Easy vector (Promega Corp.), following the manufacturer's instructions. The ligation mix was then electroporated with 10 μl of competent E. coli cells. Cells were then mixed with 500 μl of SOC medium, incubated at 37°C for an hour, and then 50 μl of the cell culture was plated onto X-gal/Amp/IPTG/LB plates. Following overnight incubation at 37°C, five white colonies and one blue colony were picked from the plate into 2 ml of LB broth. After overnight incubation at 37°C, half the culture was used to prepare glycerol stocks, and the other 1 ml was used for the plasmid extraction. Plasmid DNA was purified using a QIAprep Spin Miniprep Kit (QIAGEN Inc.) following the manufacturer's instructions. The extracted plasmid DNA was subjected to agarose gel electrophoresis to confirm the cloning of the PCR product and estimate plasmid concentration for sequencing.
Sequencing of C. krusei enolase
The plasmid was then used in a sequencing PCR reaction, where reaction solutions containing 5 μl of the plasmid DNA, 1 μl of sequencing primer, 1 μl of 5x sequencing buffer, 2 μl of terminator dye and 1 μl of PCR water. The cycling parameters were 96°C for 6 min: 24 cycles of denaturation for 10 sec at 95°C, annealing for 5 sec at 55°C, an extension for 4 min at 60°C, with a final extension of 5 min. The sequencing reaction products were purified by ethanol precipitation and sent for sequencing on an ABI sequencer. A second set of internal sequencing primers was synthesised using the initial sequence data, in order to obtain the complete sequence of the PCR product in both directions.
Nucleotide and protein sequence accession number
The partial nucleotide sequence of the C. krusei gene and its annotation has been submitted to GenBank and assigned accession number EU293890.
The following enolase amino acid sequences, including enolase 1, 2, α, β and γ isoforms, from different organisms were obtained from the GenBank and SwissProt databases. α-Enolases (vertebrates): Xenopus laevis, P08734; Alligator mississippiensis, Q9PVK2; Anas platyrhynchos, P19140; Gallus gallus, P51913; Sceloporus undulatus, Q9W7L2; Python regius, Q9W7L0; Mus musculus, P17182; Rattus norvegicus, P04764; Macaca fascicularis, Q4R5L2; Homo sapiens, P06733; Pongo pygmaeus, Q5R6Y1; Bos taurus, Q9XSJ4; Trachemys scripta, Q9W7L1. β-Enolases (vertebrates): Gallus gallus, P07322; Oryctolagus cuniculus, P25704; Mus musculus, P21550; Rattus norvegicus, P15429; Homo sapiens, P13929; Bos taurus, Q3ZC09; Sus scrofa, Q1KYT0. γ-Enolases (vertebrates): Gallus gallus, O57391; Mus musculus, P17183; Rattus norvegicus, P07323; Homo sapiens, P09104. Invertebrate enolases: Penaeus monodon, O96656; Homarus gammarus, P56252; Drosophila pseudoobscura, O44100; Drosophila subobscura, O44101; Drosophila melanogaster, P15007; Loligo pealeii, O02654; Caenorhabditis elegans, Q27527; Schistosoma japonicum, P33676. Plant enolases: Chlamydomonas reinhardtii, P31683; Solanum lycopersicum, P26300; Mesembryanthemum crystallinum, Q43130; Spinacia oleracea, Q9LEE0; Lupinus luteus, Q9M434; Alnus glutinosa, Q43321; Ricinus communis, P42896; Hevea brasiliensis, Q9LEI9; Hevea brasiliensis, Q9LEJ0; Arabidopsis thaliana, P25696; Oryza sativa, NP_001064223; Zea mays, P42895; Zea mays, P26301. Fungal enolases: Curvularia lunata, Q96VP4; Alternaria alternata, Q9HDT3; Aspergillus oryzae, Q12560; Candida albicans, P30575; Candida glabrata, Q6FQY4; Candida glabrata, Q6FTW6; Debaryomyces hansenii, Q6BI20; Debaryomyces hansenii, Q6BTB1; Ashbya gossypii, Q756H2; Saccharomyces cerevisiae, P00924; Saccharomyces cerevisiae, P00925; Candida glycerinogenes, ABO28523. Protist enolases: Entamoeba histolytica, P51555; Mastigamoeba balamuthi, Q9U615; Plasmodium falciparum, Q27727; Toxoplasma gondii, Q9BPL7; Toxoplasma gondii, Q9UAE6. Archaebacteria enolases: Methanococcus jannaschii, Q60173; Pyrococcus abyssi, Q9UXZ0; Aeropyrum pernix, Q9Y927; Pyrococcus horikoshii, O59605. Eubacteria enolases: Mycoplasma genitalium, P47647; Mycoplasma pneumoniae, P75189; Ureaplasma parvum, Q9PQV9; Escherichia coli, P0A6P9; Haemophilus influenzae, P43806; Helicobacter pylori, P48285; Aquifex aeolicus, O66778; Streptococcus thermophilus, O52191; Thermotoga maritima, P42848; Treponema pallidum, P74934; Archaeoglobus fulgidus, O29133; Haloarcula marismortui, P29201; Mycobacterium leprae, Q9CD42; Mycobacterium tuberculosis, P96377; Deinococcus radiodurans, Q9RR60; Staphylococcus aureus, O69174; Pseudomonas putida, Q88MF9; Chlamydia muridarum, Q9PJF3; Chlamydia trachomatis, Q3KLB0; Chlamydia pneumoniae, Q9Z7A6; Zymomonas mobilis, P33675; Nitrosomonas europaea, O85348; Bacillus subtilis, P37869.
CLUSTALX  was used to generate a multiple sequence alignment of all enolases. Once aligned, the MEGA4 program  was used to create a Neighbour Joining Tree in order to show the evolutionary relationship between the amino acid sequences. A Poisson correction distance and a gamma distance (α = 1.05) were used. Bootstrap probability  with 2000 replications  was introduced to assess the statistical significance of groups in the phylogenetic trees.
The alpha enolase protein sequence from C. krusei was modeled using the ModBase  homology modeling server with a psi-Blast E-value threshold of 0.01. The coordinates of the crystal structure of enolase from S. cerevisiae (PDB structure 2AL1, crystallized at a resolution of 1.5 Å) in complex with phosphopenolpyruvate (PEP), 2-phospho-D-glyceric acid (PAG) and Mg2+ was used as a template to model the query sequence. The resulting C. krusei enolase homology model was examined for overall sequence and structural differences and, in particular, in and around the active site, in order to try to identify distinct amino acid side chains that could selectively bind a ligand that can inhibit the function of this fungal enolase compared to other human and yeast enolases.
The typical fungal enolase gene coding sequence is 1,300bp long and encodes a 440 amino acid polypeptide. Excluding the primer sequences, the amplified PCR product contained 1264 bp of the C. krusei enolase gene coding sequence, from which a 353 amino acid sequence was deduced (see Fig. 1). This corresponds to 92% of the complete C. krusei gene coding sequence. The C. krusei enolase gene was interrupted by introns. Such interruption of the enolase gene by an intron was first reported in Neocallimastix frontalis .
Click on the thumbnail to enlarge the picture
|Figure 1: Nucleotide sequence of the enolase gene of C. krusei. The nucleotide sequence was determined from cloned PCR amplified C. krusei enolase DNA using the degenerate primers EnoFor1 and EnoRev2. Underlining at the 5' and 3' termini of the sequence indicate the EnoFor1 and EnoRev2 primer sequences, respectively.|
Sequence analysis of enolases
Enolase proteins are observed as highly conserved fungal allergens . The phylogenetic analysis of all enolase genes points to early gene duplication events . The homologous enolase sequences in the archaebacterial genome were designated as 'enolase-1' and the distantly related as 'enolase-2' . Phylogenetic analyses revealed that two gene duplication events occurred in vertebrates. This led to the subsequent three isozymes α, β and γ depending on their tissue-specific functions .
Phylogenetic tree analyses of the fungal enolase amino acid sequences with those of a selection of mammals, fish, plants and bacterial enolases show that the fungal sequences form a distinct clade. Fig. 2 shows the grouping of the fungal enolases together, the grouping of plant enolases together and the grouping of enolases of vertebrates and other organisms together in separate groups. The phylogenetic tree of enolase amino acid sequences show that the C. krusei enolase assembles on the tree with the fungal genes.
Click on the thumbnail to enlarge the picture
|Figure 2: Phylogenetic tree of amino acid sequences of enolases using the neighbor-joining method. The prefixes A, B and G preceding taxon labels in vertebrate enolase sequences indicate α, β and γ isoforms, respectively. Numbers 1 and 2 following taxon labels indicate enolase 1 and enolase 2 genes, respectively. The tree is rooted with eubacteria.|
C. krusei enolase shares high homology with C. glycerinogenes and C. albicans enolases but lacks conservation in the substrate binding region 137-167 and 324- 348 (S. cerevisiae numbering) as shown in Fig. 3. The human enolases α, β and γ share very high identity with each other in the substrate and metal binding sites (Fig. 3). Therefore, we refer to only human α-enolase for further description.
Click on the thumbnail to enlarge the picture
|Figure 3: Multiple sequence alignment of C. krusei enolase with C. glycerionogenes, C. albicans, S. cerevisiae and human α-enolase. The amino acids in the multiple sequence alignment are colored according to their physicochemical properties. The numbering is according to S. cerevisiae enolase.|
The N- (residues 1 to 134) and C-terminal (residues 143 onwards) domains of enolases exhibit various degrees of conservation in particular taxonomic groups with sequence identity ranging from 22% to 99%. The sequence similarity in fungal species in both domains is around 57%. High sequence variation is found in the regions 140-145 and 310-320 in the fungal species whereas this region is highly conserved in vertebrates. A. alternata and A. oryzae show presence of two additional residues Glu-Lys and Val-Lys at position 259 and 260 respectively. Comparison of invertebrates with the fungal species showed higher variation in the C-terminal domain of enolase. All invertebrates showed the presence of insert residues at position 259 and 260. The N-terminal domain a fragment with the sequence Asp-Ser-Arg-Gly-Asn-Pro-Thr-Val-Glu (approx. from residue 15 to 23), or its slightly altered form found in most of the studied enolases, is not present in enolases from A. fulgidus, D. pseudoobscura, D. subobscura, C. reinhardtii and P. putida. The Glu-Trp-Gly-Trp-Cys-Lys insert is present exclusively in enolases from plants, P. falciparum and T. gondii. It has to be noted that the enolase protein from the protist P. falciparum does not assemble on the tree with fungal genes whereas the enolase nucleotide sequence from P. falciparum is reported previously to cluster with the fungal species . In general, our results are analogous with the comparison of protein enolases reported previously .
Analysis of the plasminogen binding site
Cell surface α-enolase acts as a receptor of plasminogen on both eukaryotic and prokaryotic cells . A 45-kDa protein with strong plasminogen-binding activity has been identified on the surface of S. pyogenes and has been shown to have α-enolase activity . α-Enolase binding to plasminogen in S. pneumoniae and S. pyogenes is reported to induce the transformation of plasminogen into plasmin, and plays an important role in tissue invasion by preventing the generation of fibrin clots. The crystal structure of α-enolase from S. pneumoniae reveals the presence of two plasminogen binding sites . The first binding site contains C-terminal lysine residues whereas the second binding site is described as the internal motif responsible mainly for plasminogen binding. Both binding sites are present in S. pneumoniae and S. pyogenes. However, enolases such as those from C. albicans , Aeromonas hydrophila , Onchocerca volvulu , Fasciola hepatica  and Leishmania mexicana  lack the C-terminal binding site. Hence there have been conflicting reports about the classification of Candida albicansenolase as a plasminogen and plasmin binding protein due to the lack of lysine residues in its C-terminal [22, 27].
The internal motif usually contains both positively and negatively charged amino acids in this loop region flanked by hydrophobic residues, as in the case of S. pneumoniae, S. pyogenes and L. mexicana. Sequence analysis of C. krusei enolase indicates the presence of a plasminogen binding motif (Tab. 1), but it lacks the negatively charged residues in the internal motif, as is also observed in the case of human  and Pneumocystis carinii enolases .
|Table 1:||Comparison of plasminogen binding sites for representative enolases.|
Homology model of C. krusei enolase
The 3-D homology model of C. krusei (Fig. 4a) clearly exhibits the shape and structural characteristics of the respective template structure upon which it was modelled. Amino acid residue changes in the enolase active site could potentially be useful in the design of an inhibitory drug. Tab. 2 lists the important residues that make up the metal binding and active sites in C. krusei, C. albicans, S. cerevisiae and human.
Click on the thumbnail to enlarge the picture
|Figure 4: Representation of the active site in the homology model of C. krusei. (A) The homology model of C. krusei enolase represented in secondary structure. The coordinates for the high resolution crystal structure of enolase from S. cerevisiae (PDB structure 2AL1) were used as template for the model of C. krusei. (B) Substrate and metal binding sites in C. krusei model. The metal ion and the active site residues mentioned in Tab. 2 are represented by sticks. C. krusei enolase lacks Leu343, Lys345, His159 and Gln167 (S. cerevisiae numbering), highlighted in magenta, which form the substrate binding site in α-enolase of S. cerevisiae, human and C. albicans. The Mg2+ ion coordinated to Ser39 is shown as a dark green sphere and the other Mg2+ is shown as a light green sphere. The substrates PAG (in orange) and PEP (in blue) are shown as balls and sticks.|
|Table 2: Residue information for the binding site of enolases.|
|Feature||Human||C. albicans||C. krusei||S. cerevisiae|
PEP to PAG
|Location of amino acids in the binding site of C. albicans, C. krusei and human enolases based on the crystal structure of S. cerevisiae (PDB structure 2AL1) and a multiple sequence alignment of enolases. Data were obtained from PDBSUM.|
Analysis of the metal binding site
Magnesium is the natural cofactor of enolase. Enolase has its highest activity in the presence of Mg2+ ions and hence is known as a metal-ion-activated enzyme . Enolase contains two Mg2+ ions which contribute to the catalysis. The first magnesium ion induces a conformational change in the active site and enables binding of a substrate molecule, followed by binding of another magnesium ion to complete the catalytic reaction. The metal ion binding site is highly conserved across human, C. albicans and C. krusei (Tab. 2), as revealed by multiple sequence alignments of enolases and a model based on the crystal structure of S. cerevisiae.
Analysis of the substrate binding site
α-Enolase is an enzyme that catalyses the dehydration of 2-phosphoglycerate (PAG) to phosphoenolpyruvate (PEP) . The carboxyl group present in the substrate forms hydrogen bonds with Asp246, Glu295, Lys396 and Asp320. The phosphate group of the substrates establishes ionic interactions with Arg374, Ser375, His159 and Lys 345. His159 also forms an intramolecular main chain hydrogen bond with Asp210. These interactions were extrapolated from the 2AL1 crystal structure, as described in PDBSUM. Superimposition of the amino acid residues in the active site of the C. krusei model and S. cerevisiae enolases reveals differences in the substrate binding site (Fig. 4b). C. krusei lacks Leu343, Lys345, His159, and Gln167, which form the substrate binding site in alpha enolase of S. cerevisiae, human and C. albicans (Tab. 2). The deletion of these residues in the active site results in a larger cleft that may accommodate a larger selective inhibitory molecule that would not fit in the active site of human enolase.
The biochemical characterization of the enzymatic reaction catalyzed by C. krusei enolase has not been reported. Nonetheless, some predictions for substrate specificity can be drawn on the basis of sequence and structural comparisons with different enolases. The yeast enolase crystal structure shows the presence of two magnesium ions. The N-terminal loop is known to close on the active site via chelation of Ser39 to the Mg2+ metal ion in yeast enolase, whereas a single Mn2+ ion is seen ligated to three carboxylates and three water molecules in lobster enolase . Conformational changes in loops containing Ser39 and His159 contribute significantly to binding of substrates in the active site of both yeast and lobster enolase. Ser39 is highly conserved in the species discussed in Tab. 2; however, His159 is not present in C. krusei.
Enolase catalyzes the reversible dehydration of 2-PGA (2-phospho-D-glycerate) to give PEG during glycolysis in a step wise manner. In the β-elimination reaction, an OH− is eliminated from the C3 of an enolate intermediate. This intermediate is formed upon removal of a proton from C2 of 2-PGA by a base in the active site. Site directed mutagenesis has indicated which residues are involved in acid/base catalysis in yeast enolase. Mutations of any of the four residues Glu168, Glu211, Lys345 and Lys396 in the active site of yeast enolase were found to lower the activity in the overall reaction as compared to wild type yeast enolase by a factor of 104-105 . These residues are conserved in human and C. albicans but C. krusei lacks Lys345.
In the absence of biological data, we can only speculate that C. krusei enolase might have diverged to recognise other substrates and catalyze different reactions, or that it may compete with human enolase substrate upon infection. It would be expected that C. krusei enolase ligates two divalent metal cations based on the conservation of the residues in the metal ion binding site. A comprehensive library screening of members known to catalyze different reactions  could aid in exploring the specificity for C. krusei enolase by in silico docking.
Enolase has been previously studied for its possible use as a drug discovery target. Hannaert et al.  investigated enolase as a drug target against Trypanosoma brucei, a protozoan organism responsible for sleeping sickness in humans. In that investigation three key amino acids were highlighted as possible targets for selective ligand binding. Although there is a high level of conservation, there are significant amino acid residue differences between C. krusei and human enolases within the active site cleft. Whilst X-ray crystallographic determinations would validate our structural predictions, it is possible to conclude that there appears to be fungal specific amino acid residues differences that may be exploited to design a ligand to selectively block the active site of C. krusei enolase.
The sequence of C. krusei enolase has been determined. We have predicted the plasminogen binding site in this enzyme. Importantly, we have also identified the absence of specific residues in the active site in C. krusei enolase, which may enable the use of enolase as a molecular target to be exploited for the design of new anti-fungal drugs to selectively block this enzyme.
We would like to thank Joseph Qin and the team at Rockeby Biomed and Assoc. Professor David Groth at Curtin University of Technology for their much appreciated knowledge, assistance and resources.